Ask an Expert – July 2017

Welcome to the first edition of ‘Ask an Expert!’

At TorontoAdOps we believe in the saying “a rising tide lifts all boats.” Through our anonymous survey system (https://www.surveymonkey.com/r/TTPKC3Y) we have collected the most burning, intriguing and bullsh!t cutting questions of direct and general nature as it pertains to digital marketing and key associated topics. These questions ranged from career advice to shedding light on some of industry’s current practices. Our group of experts has anonymously provided their respective and individual opinions and points of view for the purpose of enlightenment, education and empowerment for all stakeholders within the digital marketing ecosystem.

Expert-provided opinions are based on each expert’s particular trade, level of expertise, past or present experiences and situational exposure. Toronto AdOps’ experts is currently a group of 10+ vetted, highly skilled and experienced individuals with roles and positions ranging from manager to SVP levels across publisher, trade desk, agency, brand, tech platform and regulatory organizations.

We present to you the inaugural edition of “Ask an Expert!”

**Please note expert opinions and points of view are not necessarily cold, hard facts but rather are perspectives and informed opinions and should only serve as a guide in your own decision making.**

—–Ad Fraud & Verification—–

Q1:How can I objectively evaluate a digital vendor to ensure that there isn’t an funny business going on with their technology/inventory? Are there questions I can ask, data I can see, or people I can talk to about a particular vendor? Too often I hear from my programmatic team that they prefer not to work with a particular vendor yet their reasons for not wanting to are never put down on paper, and while I believe them, they can never show me with hard evidence. Vendors like Tube Mogul come to mind…

• Where does their data come from? How is it captured? What are their inventory sources? Pulling reports on viewability and comparing them to click rates can indicate whether the impressions are trustworthy. There are so many vendors out there so personally I would also want to hear what their selling prop is – what makes them unique? Ask to run test campaigns and compare them with trusted-like vendors. In the case of some vendors, from experience I’ve found some product offerings can be shady for a lack of a better term, where a vendor can  boast to the agency about their video/display/mobile inventory and the accompanying brand studies but their results are less than lackluster.
• Ask vendors these questions: is your platform proprietary? can I get a demo? can you walk me through the campaign set up module of your platform, (a vendor who can’t/wont’ walk you through their campaign set up is huge red flag, some vendors will tell/show you their amazing dashboard – that’s not enough, (you want to see the campaign set up and targeting flow) where does your data come from (answer must be super specific, it can’t be ‘data comes from our partners’), how is data collected and protected. how are you different from your main competitors, how big is your team in Toronto (or whatever city you are in), what 3rd party verification partners do you work with. Reasons why your trade desk doesn’t want to work with a vendor could be – poor viewability stats, poor brand safety results, poor inventory quality, poor customer service, lack of differentiation from the platforms trade desk is already using. Keep in mind, a lot of vendors’ platforms are build to produce yield (buy low/sell high) so you end up with subpar inventory quality. Your trade desk team might already be aware of which vendor platforms operate this way. Though it is something that’s hard to put on paper. Keep trusting them though they sounds like they know their stuff.
• Use third party verification to separate a media seller from media measurement. Don’t buy media on a “non-disclosed” basis. Disclosure means knowing the cost of the media, the cost of the technology and the cost of the data. In addition Disclosure means knowing what the service fee from the trading desk or agency is. If an agency programmatic team doesn’t want to use a specific vendor, the answer maybe because the programmatic team wants to keep things “un-disclosed” so that the service and tech fees are bundled with data and media costs. That could result in a higher margin for the programmatic team. The opposite is true as well, a programmatic team may be pushing complete transparency and eliminate vendors that are using Non-Disclosed deals as a way of hiding :”funny business”.
• This answer will vary depending on the type of vendor/technology you will be evaluating. I would also press your programmatic team for complete transparent answers, too often opinions are given on faulty understandings of the vendor offerings or in the agency world because of some soured personal relationship. DSP (includes the likes of FB, Twitter, Snap, etc) – Only work with self serve capable platforms, if you can’t log into the system and verify they are targeting/delivering what they claim to be then you may very well be getting hosed. This can function in a managed service capacity as well, a vendor can create an account for you and provide you with read only logins for your perusal while they manage your campaign. You should be able to pull reports in platform and verify you are getting what you requested along with double checking targeting parameters. This will also give you insight into other potential offerings each vendor can provide. Network, Direct Publisher – First off it’s probably not a good idea to run with a network in the first place unless it’s something reputable (think Rogers, Bell, Corus type properties) or a GDN type of solution. But if your going to dive in this pool I would tack on some ad tags that can pull you domain level information and maybe some Double Verify pre bid blocking type of solution to avoid any potential tomfoolery on the vendor side. DV may be slight overkill for a solution but it’s like having a conservative grandmother overseeing every domain you could potentially serve on and blocking anything she doesn’t agree with (usually a lot). You can customize what you want blocked or specify by domain if the network is transparent (Ha!). 
• We are fortunate enough to work in a smaller industry. Reaching out to peers or ex-colleagues is one way to discover information about businesses. On the buy side, if you suspect a vendor isn’t telling the whole truth, then look at the performance metrics. Lots of impressions, and lots of clicks, with very little conversions, should be a flag. Or lots of impressions with no clicks. There are other metrics you could look at depending on the type of inventory you’re buying. Video will have a different set compared to display or search. Let the numbers help tell the story, if you want to be objective.
• Objectively this is difficult. You can ask a platform if they work buy side, sell side, or both. For those working both, there are often lingering questions about impartiality that prove impossible to shake off. Transparency is key, the number one question for me would be, how does this digital vendor make money and can they present their numbers laid bare? If not, if there is something to hide somewhere, then non-disclosable margins are likely being taken.
• The evaluation of vendors is depended on the KPIs and nature of the programmatic execution. Our recommendation is to connect with the vendor yourself and ask them for hard facts on measurements that are important to the business. The programmatic team should supply you with the POV on the vendor or at least a business explanation on where substitutions can be found under their way of doing business. If they do not then ask about the main KPIs with which the vendor selection happens and then inquire with the specific vendor. Keep in mind that there may be contractual obligation at an executive level which you are not privy to.
• On the publisher side I probably get about 15 emails a week from different demand partners that want to work with us. I have a standard list of questions that I ask if I don’t know anything about them. Additionally I maintain a network of other publishers who i reach out to to get feedback around what their experiences have been (payment problems, over promising etc).The basic question I use are as follows: 1. On which platforms do you have the most demand? 2. Are you able to provide SSL Demand? 3. What exchanges do you normally transact in? 4. Are you able to transact via a Private deal? 5. What percentage of your demand is unique demand? 6. Which bidder do you use? 7. What ad formats do you specialize in? 8. In which geo’s do you have the most demand? 9. What Publishers do you currently work with? Are you able to provide references ? 

Q2:Is third party verification pixels a client overstepping their bounds or being savvy with their money? Also, what is the technical overhead of having a third party verification or blocking pixel running on the tag?

• It’s a little of both but more savvy. THey want to ensure human eyeballs are seeing the ad and not bots. The technical overhead depends on how you are trafficking the creative. Sometimes you can append the pixel on original ad server. Other times it’s implemented through an SDK for mobile apps. It all depends on the setup for your campaign
• Savvy – it’s the way to measure quality being received by the advertiser. How else are they supposed to know? Rely on the companies selling them media that it is all good?
• client being very savvy. too much shady stuff out there to not keep track of your media impressions.
• Savvy as hell, trust is a rare commodity in this industry as it should be. Too many vendors sell the cart before the horse and force Ops teams to become ‘too creative’ or just straight up have to go rogue to get something to deliver or perform. This 3rd party verification mitigates this type of behavior and is a plus from the client perspective, as an Ops member when I see 3rd party verification pixels it forces me to cross my t’s and dot my i’s. The overhead for verification pixels is usually just an added CPM fee similar to ad serving, however pre bid blocking is another story as it inflates your served CPM substantially and will cause pacing problems if you’re seeing a high block rate. Meaning 50% of your ads are being blocked so the vendor will have to pace off of your 3rd party verification numbers, downside is your ad server will register the ads blocked and charge you for them as well so it will make the agencies life harder as well.
• One 3rd party verification pixel can make sense, as long as the client team is aware that no 3rd party verification service is going to return an “all clear no problems here” result. If those verification services don’t find issues, then they’re out of a job. One pixel properly trafficked should be reasonably low impact and low cost. Seven verification pixels will start causing latency and ad delivery problems, while also eating into media budgets. If verification is desired, work with one trusted partner and review periodically.
• I think that advertisers will always try to get as much data to understand the value of what their ad spend is bringing to them and I think that ultimately, transparency is important. Transparency needs to work both ways though

Q3: which is better – trusting 3rd party brand safety tools – or – managing a blacklist – or – a combo of all 3.

• There is never such thing as being too safe. Would definitely recommend implementing 3rd party brand safety tools (pre and post bid if possible) and managing a blacklist.
• I would say implement a Whitelist as much as you can
• Whitelists, only run on trusted Exchanges that are doing a good job at kicking out fake sites, domain spoofers and other fraud on the SSP level. Use third party verification and brand safety tools too.
• Have a whitelist. But even good publishers have stories you would rather not be on, so use Brand safety on a page level as well. Use a third party auditor to verify the quality of the media. Black listing is whack-a-mole
• Combination for sure, if you’re not leveraging some sort of 1st or 3rd party brand safety tool, blacklist, platform block list (blocking known data centres and botnet IPs) and depending on your environment some sort of –ve KW list you might find yourself in some hot water from a delivery standpoint plus have a real tough time when that question comes along of ‘what are you doing to keep our brand safe’ because rest assured they will ask, and ask repeatedly depending on their knowledge level.
• Do everything you can. Make sure you are using pre-bid safety tools as well as post-bid verification tools.
• I would always recommend doing your due diligence. If you work with a partner that you trust, there is no reason you can’t also manager rules/blacklists. Obviously, this also depends on how involved you want to be. But, I err on the side of caution, and ensure that if I have the ability to have input, then I exercise it.
• If buying heavily on the exchanges, even more than those 3. A regularly maintained blacklist, platform brand safety tools on, not buying mature, or unknown exchange inventory. Full sensitive category block on, IAS 3rd party anti-fraud on. Sensitive category block on, negative keyword lists in place, middle of the night daypart in place. Max it out!
• All three are necessary to ensure highest levels of brand safety. Trusting the technology to avoid unsafe environments is one of the safety pillars, the other one is manual blacklisting based of feedback and third one is in console corrections. All three elements should ensure that you are maximizing your protection.
• combination of all 3 will always be the best. Each buyer/advertiser has their own unique needs so only relying on a third party alone will give them the control that you as the buyer should be managing

—–Ad Serving—–

Q4: We have been getting requests in from Europe to serve ads here in Canada from a few different agencies. The tags we have been getting all seem to come back as malware when tested in DFP. The Agencies have been scanning them through (TheMediaTrust) and have been getting no alerts. Is this a problem anyone else has been having? Note, these are not through an ad network but direct.

• what does your google rep and DFP support say?
• Sounds shady, are you sure they are not a bad network trying to buy through an agency, infecting the ads in Canada to drop malware onto computers.
• I’ve never worked with any European advertisers serving in Canada, my thought is that if you are seeing this happen across every vendor/pub you contact then maybe there is some truth to it? Unless I had a live tag I wouldn’t be able to troubleshoot the issue.
• This happens with any third party testing. DFP and TMT use different approaches when scanning tags/creative. In the DFP malware alert, are you noticing the same URLs constantly appearing? I would suggest looping in with your DFP account team to confirm what is triggering the alert. Usually you will receive some information that you can use to respond to the partner who sent you the tags.
• Google’s market share in Europe is much lower than here, each country has historically developed and used a more proprietary ad tech stack. The tags may simply be from outside the DoubleClick tech stack and because of that, are not so compatible.
• I don’t trust anyone I’ve never heard from. A lot of 3rd parties backfill their unique demand with other exchanges so ultimately you are impacted by whatever blocks/rules that they do or do not put into place. Not a good scene

—–Career Advice—–

Q6: What’s the next move career move if you are very knowledgable in ad operation, work with tagging strategy, optimization, attribution, RTB, SEO and is client facing. Is becoming a digital marketing consultant the next move? appreciate your input on this.

• keep developing your skills. get some commerce experience, as well as ad tech/marketing tech/database platform integration experience. then it’s probably trying to get into consulting. companies that require consulting never need to be consulted on any one thing (for that they hire contract positions). most of the time a consultant is engaged when there is a need to integrate or combine a number of things that normally do not live in the same scope of activity. An example of when a consultant would be engaged is something like: connecting offline CRM database to ecommerce platform to order process/intake system to data management platform to digital media buying platform… and to make sure the flow of data and information is bi-directional.
• Go client side, quit chasing the dollars on the tech side (unless you want to get into the duopoly of FB or Google). Have a direct opinion on strategy from the ground up and bring that expertise to where it could have the most impact
• The short answer is ‘it depends’. Are you also interested in leading a team? Have you already had that kind of experience? If you are more comfortable as an individual contributor, then there are lots of companies looking for the experience you just mentioned. Sell side and buy side need people who have a well-rounded knowledge of the ecosystem. Becoming a consultant is a path, if you find yourself wanting to change roles often, and work in a variety of companies/teams. But, if you’re looking to add experience, then your next move could be the opposite side of the industry that you’ve been on.
• Client side is trending hot right now.
• Digital Marketing Consultant is definitely one of the ways to use your skills to advance. Programmatic activation on the SSP, DSP or DMP side are also good avenues to explore as the progression of your career. Another one would be solutions based service companies which is close to consultation however is not an independent practice.
  

Q7: In terms of work-life balance , which side (agency vs publisher) is “worst” or skews more towards Work than Life?

• Really depends on the company but from experience, I’ve found that agency skews more towards work than life when comparing it against publishers I’ve worked with.
• Good question. Vendor is by far much better work/life balance.  Agency is a sweatshop with high turnover rates. You get worked hard on a minimal salary
• I have heard both. Some DSPs are sweat shops for ad ops with a hard enviroment and some Agencies can be too. It’s finding a good place that tries to give you both. Whenver you work in Ad Ops you already know it’s not a 9 – 5 job, so ask around and find a place that appreciates their employees.
• agency is the worst, that may be aggressive but I’ve yet to hear a different opinion and experienced it first hand
• I’m sure you’ll find different answers from each side. In my opinion, you’ll find more balance on the supply side. That is my experience, and what I have heard from others.
• Digital media in general is pretty intense and 24/7 pub or agency side. Agencies probably work more hours, but offset that by partying down a bit more and having slightly more fun.
• This depends on how business is run. Typically agencies are more skewed towards longer hours however this may not be the case in some situations
• Agency life tends to have a repuation for being more demanding but each company is different. Do your research before accepting a role in any new company to understand what you are getting yourself into
  

Q8: What’s the hour rate for someone who know RTB, ad ops, and search?

• If you’re doing it as a consultant you would ask for more than the hourly rate assuming you work 40 hrs/week for the year. If you know the 3 pillars, your salary for a full year would be anywhere from 55-80K, which translates as 26-38.5/hr. As a consultant you are working per project so that could be a 25% on top of that
• depends on your expertise but hourly rates for a programmatic consultant with 5-7 years ranges from $200-$400 USD but I have very little hands on experience in this area at the moment
• Don’t charge per hour, charge per project. With consulting it should not be about how much each hour is worth to you, it should be about how much the end goal is worth to your client.  
• Depending on all sorts of things, $20 – $25 per hour.
  

Q9: What do you think are the top 5 skills you need to be a programmatic manager? Either soft skills or hard skills.

• Detail-oriented, tech-savvy, quick learner, organized, analytical
• curiousity, experience on all major platforms, people skills, translating complex to simple, constantly learning
• be able to talk about it using normal words. Have complete understanding of ad serving, understanding of how attribution works, understand how yield management works, be able to call out bullshit when you see it.
• Here are a few: 1. Attention to detail; 2. Analytical; 3. Organized; 4. Multi-tasker; 5. Problem solver;… and many more
• Console knowledge, Communication, Presentation, Analytics (attribution) and people management are the top 5 skills
• 1. You must be curious and love problem solving, 2. Solid excel skills (pivot tables, vlookups etc. 3. understand technology.
  

Q11: I have experience as marketing manager and expert in b2b and govermental space and keep updated myself by researching and actively reading about digital marketing but did not worked in digital/ social media marketing? What job title would be my first step to enter?

• entry level
• Digital marketing coordinator.
• Without work experience in digital, you’ll probably be starting as something such as a Media Coordinator.  
• It depends on the business entity and their needs. Account Management would be one of the first steps to take in order to understand the applications in digital or social media.
  

Q12: Is there an opportunity to disscuss my resume and abilities with an expert?

• sure.  send an email to info@torontoadops.com or join this group https://www.facebook.com/groups/OnlineJobsToronto/
  

Q13: I’m a student trying to find an internship/entry-level job in the field. How do I make agencies get back to me?

• look to the tech side for an internship, you’ll find them much more receptive and you’ll enjoy some sweet perks while working for free. Plus you’ll certainly learn more
• Be different in your cover letter and resume. Inject a bit of personality into it.  Make people remember your resume.
• Network your ass off. Find a friend, or a friend of a friend who works at an agency and get them to put your resume on their managers desk.  
• Perserverance is the key, pleasant and kind communication. Human Resource divisions are faced with 10-30% turnover rate therefore they might have their hands full on full time candidates. Communicate your willingness and check back multiple times until a response is given.

—–DSPs—–

Q14: I sat in a meeting with someone at Bell Media who was pitching about real time verified walk-in campaigns. They claimed SITO mobile’s platform has advanced to the point where they are seeing the walk-ins in real time, allowing them to optimize a campaign towards store level visits opposed to an online conversion like CTR. Curious to hear your thoughts on SITO mobile’s platform

• Who verified it? Bell? Also this is one media touchpoint not all the brands touchpoints, optimizing on a per channel basis rewards placing ads in front of consumers who were on their way to purchase anyway. Have we learned nothing from last click attribution?  
• I can’t speak to SITO specifically, but this is plausible. For example, in our DSP we do effectively the same thing through partnering with Cuebiq — their SDK is embedded in many apps, so they have a rich stream of location data for a subset of users, and overnight they process the data and send us a postback with the details needed to match to the last impression/click before the walkin, so that we can record the conversion — as though it were any other — in the DSP. Such an approach will only capture a subset of the actual walkins — but for optimization purposes, that’s enough. The end reporting should instead be based on lift vs.
  control after some number crunching post-campaign.  
• I do not know SITO’s platform or transparency into their solution but I am extremely intimate with this type of solution. I helped create one based on 1st party verified app location data this year in fact. Questions and context below to ask SITO or any other vendor offering visitation data.
  First off what do they base their location data off of (targeting and reporting)?
  99% of the time the reporting and targeting is based off of bidstream data, which is easily the most scalable, but it is riddled with inaccuracies and is generally unverifiable by any standards. App owners control the lat/long provided to the exchange, they know that they increase their yield by applying this lat/long when sending traffic to their SSP. Better still they know they increase their yield even more when adding decimal points onto that lat/long falsifying the accuracy. More often than not they’ll just reverse lookup the IP address on the impression and apply a lat/long based on that, super easy. As a tell tale sign if they are using bid stream data they will not be able to provide you dwell time (how long users spend within the location). If they are doing it right and using some sort of verified app data ask them how they increase their deterministic audience reached and what is the threshold of a statistically valid panel audience.
  
  If they are using 1st party verified data, what apps do they pull from and do those apps need location to function properly (map or weather application)?
  Sign an NDA or whatever needs to be done but at the end of the day you have to know this otherwise I would assume bidstream data. If they name off some apps do your research and even reach out to them and verify this provider is in fact using their data. FYI if an app doesn’t need to use your location to function properly then the App or Play store only pass 2-3 decimal points, if location is required then you’re looking at 4-5 points. There’s a massive difference in accuracy: https://en.wikipedia.org/wiki/Decimal_degrees
  How do you draw locations, point radius or boundary?
  Boundary traces the footprint of building pending it’s a standalone and point radius draws a circle using the middle of the address the beginning of the radius and will include people walking by and parking lots where as boundaries are as efficient as you can get. If they say boundary ask them how long it takes them to draw a list of 500 locations, if they say something ridiculous like 24 hours you know they are lying. Takes us ~72 hours minimum to get a 500 location list drawn.
  What is your definition of a visit?
  Leave this one open ended and see what happens, if the solution is legit they will be able to not only define what a visit is (how long a user has to spend in location to count as a visit) but also be able to alter the time of a defined visit depending on your campaign. Example would be a QSR visit would be defined as 5-10 minutes where as a car dealership would be 20+ minutes to properly categorize someone interested in purchasing a car.
  How do you scrub employees and general noise?
  Simple yet devastating to a thin solution, employees will inflate your visits and your average dwell time to insane levels. Hopefully they store historical data and can identify these users easily otherwise it’s not an easy task.
  There are a pile of other questions to ask but it’s a lot. If preferred I can connect with this individual and drop some knowledge. I would also read the riveting page turner below, everything you need to know about location based advertising.
  http://www.mmaglobal.com/news/mrc-issues-location-based-advertising-measurement-guidelines
  
• “ this is bullshit.  verified walk-in technology in Canada sucks, very little reach, and CRTC privacy regulations are way too strict to do this at scale.  Verified walk-ins is not something anyone should be doing with a mobile media provider, it’s a waste of money.  To do this properly you need to partner with a beacon vendor or a physical web vendor.  Tell Bell Media to stop selling you cat in a bag.
• Agreed — Beacons all the way, however, with cross device retargeting coming up I could just heavy up on retargeting and BOOM I will get tons of “beacon” fires from people who were going to come organically anyways as they were already checking out the website and had an intrest. Doesn’t prove the media is working, just means I am cookie/tagging the right devices of people who already have a high chance of coming. Be careful how you measure these results. results can always be unicorn when people find clever ways to target –

—–Measurements and KPIs—–

Q15: On the client/agency side, when coming up with CPA goals and then comparing vendor performance, are they usually including display with (cheaper) media like search and native? Like when they say “I’m getting a $100 CPA with other vendors” what are the chances that they are comparing apples to apples (e.g. standard display with standard display, search with search, etc)

• You are not alone. Getting comprehensive responses surrounding CPA can at times be a challenge.
  Things you also want to consider are:
  1. Does the other vendor have access to more first party data? If they do, their performance will be better. Make sure your pixels are in as many right places as possible.
  2. Is other vendor’s budget much bigger than yours? If it is, they will have more reach, their CPA would be lower than yours.
  Try these steps:
  1. Ask the agency for full DCM report. What you want is the DCM report ‘by site’, or ‘site report’.
  2. Ask the agency for the brief they got from the client.
  3. Ask the agency for the plan they presented to the client.
  4. Ask for the blocking chart.
  5. Ask for ‘conversion path’ report from DCM.
  If agency person asks why, tell them you want to optimize your media delivery around all other digital initiative on the plan, like social media activity, direct bookings (knowing this can help you avoid overlap), search, etc.
  Try to push your agency relationship beyond the buyer you are talking to. Learn who the account manager or director is, take that person to lunch.
• How are they de-duplicating the reach of each vendor? How are they measuring the campaign reach of multiple vendors? Perhap it would be better to look at test and control Geo markets.
• This has forever been a problem, even at the agency trading desk we had to physically sit down with agency planners and separate everything out on their blocking charts by device type (mobile and tablet vs desktop), then by medium (search, display, video, social, native, etc) to compare similar vendors from a performance standpoint. Obviously our intention was to highlight where we could jump in and capture budget but also to set proper expectations and gain direction on strategy. The truly telling part is that even though we were part of the same company the aforementioned teaching moments weren’t usually available to us. We had to really work to get time with the planners and teach them the nuances between each vendor. The above is all built on the premise that the client has a solid attribution methodology otherwise all is for naught as vendors with the largest retargeting pool and search capabilities will capture the most budget. If they do have something like DCM to understand the user’s journey through each vendor it can paint a more telling picture and the planner can make more informed decisions. That’s usually asking a whole lot from an agency side but I have seen it executed before.
• Great question. Every client is going to be different in how they measure and how they compare vendors. There’s no one definitive answer. Not to throw too much shade on the buy-side, but you will most likely get told that performance is terrible compared to others even when you are the top performer. Transparency on performance is something that not all advertisers offer. In terms of ‘apples to apples’, your guess is as good as anyone’s. You’re not alone in banging your head against the wall. – C.Quinn
• No, usually on the agency side we’re not mixing CPA’s between different media formats.
• Chances are pretty good. While CPA should be a combined metric across all digital touchpoints, most agencies are still working in siloed approach. Their definition of publishers might be skewed as there are multiple elements in digital display, video, native space. Search and social are looked at independently when comparing publishers.

Q16: Why do people loose all rationale thought and apply a different standard to Google and Facebook? It seems like either could deliver a golden turd and it would make glowing headlines followed by 1 hour presentation to C level staff.

• The reason this can happen is because beyond media buying and digital marketing, both Facebook and Google act is research and information providers to the brands and C level people at those brands. All major brands turn to Google and Facebook for things like consumer surveys, trend analysis, changes in consumer behavior, etc, etc. Because of this, Google and Facebook are seen as being credible partners (outside of digital media), they are basically seen as major consumer information hubs. To C level people this is actually far more important than digital media because C level people never talk about campaigns, their focus is on a holistic level, and a lot that info comes from Google and Facebook. This is why any campaign specific information never really gets questioned. The thinking is ‘Google/Facebook provide us with super valuable information about our consumers (which we use to make MAJOR marketing decisions), so all individual campaign pieces must be accurate’. It’s basically benefit of the doubt, the research info is amazing and credible so it is assumed everything else associated with Google/Facebook is also amazing and credible.
• People do not know the right questions to ask, or have the facts/logic needed to challenge the golden turd. What happens in these C-level meetings is you have a Google or Facebook rep use big words while highlighting the brand in a very positive way. This makes the C-level people feel good about themselves, though they might not be fully understanding the big words that Google or Facebook rep is saying. But it doesn’t matter really, you are not going to question the person who is telling a whole room of people how well of a job your marketing department has done. No one bothers to question anything, it is risky to question Google or Facebook at C-level, your C-level peers might not see you in a positive light challenging the smartest companies in the business.
• Welcome to the duopoly that is the current reality for programmatic. Having never worked at FB or Alphabet I can say that it is a full time job pitching against their solutions. There is no debating on whether or not they offer full end to end solutions that are best in class, the only advantage other vendors have is they can be more nimble. With all the above I would agree that companies glorify their solutions substantially more than anyone else but let’s face it, if you choose to go with a FB or Google advertising solution no one is going to question you. If you go for an alternative you are going to get the third degree, especially if you don’t get results.
• I feel like we apply the same standards to all our partners Google and FB included. If Adobe, or Amazon become the best place for us to invest our media dollars, then that’s where the dollars will go.
• I think that people get stary eyed when the names Google and Facebook float into conversations.

Q17: how can i calculate my actual media cost, without all the technology fees

• go for a Disclosed contract with the media buying agency who is using programmatic. Buy direct from a publisher.
• Ask for placement level reporting with data, tech and media fees broken out from each vendor and learn how to vlookup. You always run the risk of someone tampering with the reports but it’s as good as you’re gonna get unless your vendor is using DFA to serve and DBM to buy and then you can link them and pass back costs into the ad server for transparency.
• best way is to engage a third party like http://www.adfin.com/ who can analyze your impressions and tell you how much of your money actually goes to working media. If you have access to your DSP, you can just run the report that will show you tech fees, etc.
• If that information is not readily available, apply researched averages to arrive at a conclusion about media investment. Please keep in mind that programmatic is a constant and that cost of technology should be blended with cost of media as there is no way to execute programmatic advertising without technology.

—–Optimizations & Campaign Performance—–

Q18: Performance campaigns, direct-response campaigns. Outside of retargeting, what are the tactics that work best for you. Do you drive performance out of user data or out of placing ads in premium environment? Anything else?

• Contextual keyword targeting in site content works very well too  
• Overall principle is to keep media costs low and hit as much reach as you can at the beginning. Once you figure out what’s working start cutting tactics that aren’t doing well and figuring out ways to increase reach into what’s working. Analyze the data that is driving conversions and put together different tactics based on that. If you’re working with a decent reporting suite and possibly a DMP you can analyze not just the converting audience you’re driving but rather the overall pool of users that have converted on the site (pending how long you’ve had the pixel/postback on the conversion page). This should tell you more than enough and if possible run this report prior to spending a single dime. There’s a heck of a lot more nuance to this process and it’s so unique to each situation, so long as you follow the generic playbook buy low to learn and buy high when you know what works (high meaning a realistic CPM that still makes sense) you should be fine in the RTB world. If i am seeing conversions coming from a ‘premium placement’ I will look into a PMP for inventory not available on the open exchange but that’s only if the conversion rate is similar or better than what I’m seeing via open exchange.
• Depending on the quality/amount of your first party data you can make decent look a like models that can perform fairly well.  Optimization is a big thing if you follow typical consumer behavour. Optimize to things that make sense, like time of day, day of week, etc.
• 1st party data is where the party is at, it’s accurate and can be kept fresh.  
• User performance has been outperformance premium environment for awhile however it’s connected to cost of purchase and creative responsiveness from the audience. Creative progression would enable performance in some scenarios. Dynamic or rich media creative coupled with viewability standards which are high could help performance.
  

Q19: I’ve always wondered do display ads on adult websites work? I just find it hard to believe someone would click an ad while doing the dirty deed. The bounce rate must be extremely high with all the one handed accidentally clicks. I was reading up on how Eat24 targeted adult sites with its provocative ads with success. However I just don’t see the value in serving ads on adult sites. Would really like to hear your thoughts on this and maybe even start a discussion. Thanks!!

• Getting someone’s attention while they are task oriented is a media challenge that goes beyond adult sites. Think of the efficacy of showing ads while people are in the middle of booking a flight or clearing their email or trying to fix something. Probably a good area of research for eye tracking and heat maps.
• Just like any other content play not every brand is going to get the desired result from advertising on porn sites. However demand for ad space is low and supply is monstrous for these sites, which means you could buy them for nothing even without the argument that ‘it’s porn’. Whether it be a content or audience play porn sites offer a massive supply source which could do wonders for specific brands, the only issue is the public perception and social norms.
• It does work and it works well. Not for all brands of course.  There is a huge programmatic supply ecosystem for adult inventory.  Stuff does work.
• Unless your brand fits, there’s really no value here for me.
• They could work – remember that they are in a non cluttered area where a lot of advertisers choose not to buy ads on these sites. It all depends on message and perception but if the client is risky enough, it could work to their advantage.
• the real question in my mind is what kinds of audience segments are they building on the back end.
  

Q20: What’s the best way to approach media planners/buyers with something that you strongly feel will provide value to their team?

• Tell them you want to set up a call with the client so you can tell them how they can better reach their business objectives and increase ROI.  Try talking directly to the planner first but do not put all your eggs into that basket. It is not uncommon for your typical agency planner or a buyer to not be able to look beyond the vanity media KPI’s and see the bigger picture (of achieving actual marketing objectives).  Push for a call/conversation directly with the client. Be proactive with this, don’t be shy.
• Highlight the problem you solve and it’s strategic advantage in the simplest way possible. Planners/buyers knowledge spectrum variance is massive so always cater to the not so knowledgable until pressed. You can usually tell by the interest/questions in the room where you have to go with it. Lastly, shower them with free drinks, food, tickets and general gifts. They are not well compensated from their employer so help them out if you have an expense account.
• Show them examples of your competitors doing it.  
• one on one communication always works. Start by outlining the benefits of what you have to propose to them and the extended team.
  

Q21: “How do ATD’s get away with charging so much for their services and delivering crappy results?

• They use Non-disclosed contracts that mask how much the service fee is and how crapping the media quality can be.  
• Truly depends on the ATD you are talking about, I only have hands on experience with one and before I left we were sharing our media costs with the clients themselves in a lot of cases. This was mainly because we operated on dCPM’s to actually optimize to performance but I understand there are some ATD’s operating in a not so transparent manner. To answer the question the ATD is a holding company that is owned by the greater agency umbrella. Example, WPP owns both Xaxis and Group M. This makes it extremely beneficial for the umbrella company(WPP) to have Group M pump a whole bunch of budget through Xaxis instead of farming it out to outside vendors and losing a lot of margin. This is a common agency practice to buy complimentary businesses in order to extract more money from the clients they manage, they exercised a similar strategy when it came to print, radio, OOH and TV as well just not to such an egregious level. So bottom line, top down direction is use the ATD regardless of performance so we can make more money and if you don’t here’s a box for your things.
• Some ATD’s are actually good.  But the ones that get away with shitty results are because their clients have no knowledge in the matter.  Clients can not tell the difference between good results and bad results.  Most of the time when you see an agency or an ADT get away with bad performance is because their client does not know any better.  It’s worth noting that it is not in agency’s/ADT’s best interest to educate their clients.
• I can’t speak to the results of their efforts, but keep in mind that ATD’s are profit centers for agencies. They sell high and buy as cheap as possible (a good portion of the time).
• They’re under pressure for their margins and lack of transparency, but the money keeps going to the ATD desks precisely because they do not deliver crappy results.  
• No sure. It might be dictated by the company or agency which you are working for. The ATD should follow planning direction. There are vast differences in how media is bought now and audience responsiveness has been dropping so please take that into account. It’s never one thing that impacts performance so help the ATDs progress in creative, and help them understand which performance is primarily used to measure success. Remember that they don’t own the inventory and nobody wants to do a bad job, they may just be lacking perspective.
• Many advertisers don’t have the budget or sophistication to bring their own buying in house, shitty ATD results are better than paying someone who knows even less to get even worse results for you

—–Programmatic—–

Q24: What is the future for DSPs in the industry?

• DBM in still in the lead, others are very far behind  
• It’s not enough to be just a DSP.  Platforms that will continue growing overtime are the ones that have something proprietary and useful at the same time.  Think of the platforms that can: 1. Provide you with unique, quality data sets, and 2. Provide you with a unique placement, at scale, that you can not buy through any other platforms. Based on those two parameters DSP’s that are top of mind are Google/AdWords, Yahoo/BrightRoll, Amazon… and of course every DSP’s bratty cousin FaceBook.
  Best Canadian mentioning here would be StackAdapt and Addictive Mobility. We will probably see more Canadian based DSP’s come up over the next year or two.  
  
• First there will be constant consolidation, but as the world becomes more connected I feel the DSP will continue being the hub of digital advertising just with new tools and forms of reaching users.
• Consolidation, there are too many right now. I suspect we continue to see fewer, but larger DSP’s.  
• In the near future, consolidation of purchase by advertisers. As we go further out, there may be a consolidation for DSP and DMP at each client level and then possible expansion to single stack end to end execution with the emergence of meta dsps. That is on the technology side. With the service component, our belief is that serice area will grow in the form of third party consultancy or execution service.
• I think that over time there will continue to be more consolidation in the ad tech industry.
  

Q25: Why does the abomination of header bidding still exists? It is plaguing the internet, making at to load much slower and decreasing brands $$ investments.

• header bidding is a balance/equilibrium mechanism.  when RTB first became a thing we saw a swing in CPM’s we were paying.  We went from $20cpm (on IO’) to $2cpm (open exchange).  When a swing like that happens economics of the situation loses balance – price of goods does not reflect the demand volume. Over the last few years we have seen the CPM’s move closer to a new equilibrium point.  First with deal id’s and PMP’s and now with header bidding.  If we talk about ‘value of the budget’ we are being short sighted.  RTB economics and yield management are a much bigger picture.  if digital inventory costs too much maybe reallocate the budget to another media.
• It’s the industries way of flipping the bird to Googles tech stack and the former ‘first look’ DBM users got at Adx. Also it makes sense but only for the publishers with insane scale that could dial into all major DSP’s to save on tech tax of an SSP and make the herculean effort to retag all of its pages.
• Not sure where you’re getting this insight. Header bidding exists to create competition. In terms of slow load times, that isn’t the case for a lot of publishers. The timeout for waiting on responses to bid requests in controlled at the publisher level, so it can be adjusted based on how quickly responses are happening. In terms of brand investment, it is having a positive impact. Brands can now compete on an even playing field across multiple platforms (not just the one).
• header bidding will go away the day that publishers can make exactly the same revenue without it. Smart pubs are able to effectively analyze the revenue generated on their site and balance that with the impact to load time on their sites. Keep in mind that the tech is always changing but at the end of the day revenue is driving these decisions.
  

Q26: Why are advertisers doing RTB themselves instead of with an agency?

• If a brand has a large enough digital marketing budget, and a strategy with omni-channel business objectives, it is easier and faster to brief, plan, execute and make quick adjustments in mid flight if you are handling it internally. whole other side to it. When you are doing something yourself (like the RTB) you can tie all other digital ‘things’ into it, that the agency would simply not be able to do (due to lack of expertise). For example, connecting your email marketing to RTB, connecting your content strategy to RTB, connecting your customer management database to RTB, connecting your brand’s mobile app usage to RTB. For example, all major brands have very powerful (often power by IBM) propensity model algorithms that they deploy across all their direct marketing efforts (call center, email, mobile app usage, POS, etc, etc) – this is not something an agency would know how to integrate with RTB (because media agencies do not focus on enterprise level data models and systems), but if RTB is done by the advertiser then propensity models can be easily integrated with RTB platforms. It is MUCH easier to achieve omni-channel marketing model by executing functions in-house than having your execution spread out between several external partners. RTB is just one small element in an advertiser’s omni-channel marketing model. It doesn’t stop with RTB, this applies to any marketing function that can be execute electronically. Email marketing, eCommerce, content marketing, digital OOH, social media, SEM.  Even some creative functions.  The more functions a brand is handling in-house the easier and faster it is to connect those functions to work in sync, and be able to have an omni-channel approach and  clear attribution look.  
  
• When an advertiser does programmatic themselves they control and own all the data, can see the actual media costs, tech fees and data charges. If an agency does it, then the contract with the agency has to be super clear and solid about who owns what .  
• Far greater transparency, control over how their campaigns get run and lastly because it’s far cheaper in the long run. Agencies are only useful for discounted rates on things like radio, television, OOH and print. They offer no added value when it comes to programmatic, in fact I would say they are a detriment. Plus brands need to get smarter about this and leverage their own internal data end to end.
• Trust and control issues are at the centre of this execution. Some clients believe that they need control of the data and by notion they expand their control over media investment execution. It is not an incorrect notion, it simply depends on what you are trying to achieve long term. Clients sometimes are left with no choice but to move the execution in house due to murky agency practice.
• Budget and expertise – its hard to find really great people.
  

Q27: What would you say are the main differences between the programmatic industry in Toronto, and the industries in London and New York City? Are the markets/industries smaller or bigger?

• In Toronto we are in a bubble situation, we have way too many suppliers than there is demand.  In New York the supply and demand for RTB are much closer together.  Over the next 2-3 years we will see a number of companies either be bought or go out of business in Toronto.
• We are seeing huge consolidation and this will continue.  
• An average Canadian IO at a tech vendor or ATD would range around $30 – $50k, in the US were talking $100K – $150. Plus the million dollar IO is a unicorn in Canada, in the US it’s called a good week. On top of that the salaries in the US are staggering in comparison to Canada.
• Compared to NYC and London, the Toronto industry is smaller. That’s the reality of being in a much smaller market. However, my company has teams in each of those markets, and I can say that our size does not reflect our abilities or our knowledge. Toronto is one of the leading programmatic markets globally. The biggest difference between Toronto and those other markets is the amount of competition that buyers and publishers go up against. In Toronto, there are fewer companies to compete against.
• NY and London are definitely bigger. Their advancement in programmatic is very much the same as Toronto however these two cities sometimes represent a centralized hub of excellence for some clients. NY is used for centralized execution across North America in some cases and London is very much the same for Europe. That is what makes them bigger as it’s consolidated investment. It is not necessarily more advanced as the common ground has to be set for multiple countries.
• from my experience the US is considerably more sophisticated than Canada. Many of the buyers in London also buy across Europe and the Asia pacific regions as well so their buying experiences/needs are different. Lastly. The adtech ecosystem seems to be more competitive in Europe.
  

Q28: Is DBM really better than the GDN?

• Yes, the tools in DBM allow you to buy cleaner, more viewable inventory and execute more high level tactics. GDN for Ad Words are meant for small buys from a do it yourself “Mom & Pop shop”. DBM is meant for big campaign budgets.   Two different technologies and both have their strategic advantages. GDN being a network has access to properties that aren’t really available on RTB, albeit mostly longtail it’s stil super useful. Plus the CPM’s were insanely cheaper when I used it circa several years ago. DBM accesses different inventory and a whole lot of it plus the tools within DBM are pretty badass from both a targeting and reporting standpoint. “Really depends on how complex your executions are and what your objectives are. With DBM you get reach and platform connectivity, but with AdWords you get creative options and media buying pricing options that are not available in DBM. Smart buyers use both platforms, how widely you use one or the other really depends on the scale of your operation.
• DBM is an enterprise solution, it can connect to many other things for a variety of purposes.  If your digital marketing operation is complex and there are other connected platforms at play (DMP, dashboards, centralized ad server, attribution models, ecommerce, email server, dynamic creative, etc) then DBM is probably the platform you will use the most.
• AdWords is a SMB solution.  It doesn’t really connect to anything else outside of a handful other Google products. If your digital marketing budget is not large, and you do not have much of an omni-channel operation then you should probably maximize the use/value of AdWords before using DBM.
• If you are using AdWords, I would recommend using DBM only for things that AdWords can not do, or if you steady find yourself underspending in AdWords.
• One thing to keep in mind, AdWords tech fee is 4%. On average, DBM tech fee is about 3.5x higher.
•  Yes   It’s a matter or perspective and the short answer from our perspective is yes. DBM is the progression from GDN with multiple advancements in place.

Q29: Can you provide an average range of a typical budget used for a programmatic campaign? Say, for a Fortune 5000 company on a new product launch.

• Take the size of the target audience multiply by the frequency you would like to hit them with. Use viewable and valid impressions. Choose an a size that makes sense for what you need to do. Find a reasonable cost for that ad size from good quality publishers that you have heard of and multiply it by the number of valid and viewable impressions you need while including how much waste (off target) will occur. That’s your ballpark media cost. Add programmatic fees on top of that.  
• What is Fortune 5000? Google returned no results for that.  Is that fastest growing companies or biggest companies? Campaign budget can vary significantly, depending on the length of the campaign as well s the type of business a company is into, etc.  Overall ANNUAL programatic budget for a large Canadian brand ranges anywhere from $3 million to $20 million. For a new product lunch you can see budgets of up to $1.5 million for a single campaign.
• The range varies by line of business, by complexity of business or simply by budget assigned to a campaign. From $10,000 to $2,000,000 which really takes all kinds of campaigns into the equation of investment.
  

Q30: Does header bidding positively or negatively impact digital media (from an advertiser’s POV).

• It will cost the Advertiser more, however, publishers need to make money to survive and produce good content  
• From an advertisers perspective it means higher yield, but god damn it is painful setting up.
• It creates an even playing field for all platforms. I would think that would be a positive impact for advertisers who work with non-Google vendors.
• It increases our CPM’s slightly, but overall has had little impact on day to day business  
• Positively, server to server integrations are needed to establish and faster, better and closer business relationship with key entities.
  

Q31: ad frequency control … where are we with this?

• Mobile POV: Almost impossible to maintain programmatically as freq cap gets controlled at app and browser level ie. facebook browser, app, chrome, safari, etc.
• – it’s a thing, meaning it’s possible so long as you’re buying through a single DSP. Once you start incorporating several vendors into a plan your freq control is gone. However, even if you use a single vendor you would be hard pressed in controlling the total ads a user receives on their phone via app and web as the identifiers are different and not usually linked.
• it is pretty weak right now.  Still same frequency control as we had 15 years ago.
• Some implement it, and some do not. Because this is controlled at the campaign level, there is no industry-wide rule. If you mean universal frequency capping per user, per platform, I’d say we’re getting quite close.
• If you mean universal frequency capping across all media formats TV, radio, OOH, digital etc. We are very far away from that.  
• We would say it’s work in progress. The industry still has a long way to go to truly understand frequency and impact of creative. Better controls have been put in place in consoles however the consoles are not interlinked therefore frequency is multiplied. We need to be aware of industry’s walled gardens and disconnects that are in place in order to use frequency of advertising to improve of audience experience with the brand
  

Q32: Why does my programmatic buying team optimize to media metrics and not to marketing or business outcomes?

• I personally always preach business outcomes, however sometimes clients just don’t care and specify something like CTR or CPC and will hear nothing else. Even though you have their best interests in mind you have to give them what they want or lose the budget due to ‘under performance’.
• Probably because they do not have a direct line of sight into how marketing is impacting business objectives. Does your programmatic team see ‘sales’ and ‘funded accounts’ from your client’s back-end system? If not, how can they optimize to that? There needs to be a direct connection, running manual reports and sending them to your programmatic team is not going to work for this.  If you do not have direct line of sight into what’s happening with business objectives, the best thing you can do is optimize to a conversion point closes to the business objective.  For example, if your client’s business objective is to ‘increase store or brand foot traffic’ the best thing your programmatic team should be doing is optimizing to ‘store or branch locator’ conversion point.  If your programmatic team is optimizing to CTR you should fire them all right now.
• Without putting myself in the minds of the buying team, I would say that ‘media metrics’ is an easier target. Also, those metrics are standard for everyone, and get the most attention.
• It might not be possible for them to attribute towards physical sales, in which case they need to use some other metric.  
• Sometimes these two clash and there are reasons such as scorecards, KPI’s written into the contract and technology restraints that disable optimization towards business outcomes. In some cases the clients do not share business outcomes and media metrics are the next best thing.

—–Vendors—–

Q33: Why do so much publishers allow advertiser to leech data off their visitors and building segments for nothing in return with stuff like conversion pixels, research pixels etc?

• Mostly because they aren’t aware of what’s being collected, but also because if you don’t allow pixels to be placed by vendors they will be hard pressed to deliver any meaningful results. Basically it’s a catch 22, all publishers can really do is remove pixels when no longer working with a partner to restrict data leakage.
• It’s not for nothing, it’s for the CPM buyers pay for that publisher’s impression. Many people believe that once you bought an impression it becomes yours, you can do whatever you want with it, including keeping the data associated with that impression. As a buyer, I have the right to analytics, right? What I do with those analytics is a whole different story.
• Publishers don’t necessarily want this to happen. But, it happens. A lot. The reality is that advertisers have the ability to do this in a variety of ways. Not all publishers understand how this works.
• I think there’s a bit of a belief that more trackers, equals more money.
• They have to do that in order to have a business model with the agencies that represent advertisers. Advertisers sometimes dictate tough terms for continuous investment and it becomes a very tough decision. Lean in and benefit from investment or stand your ground and risk losing it.
• Pub life is hard man, not everyone understands their value.