Since launching Rockerbox Attribution Platform I’ve had conversations with thousands of marketers discussing attribution. What have I learned from these talks?
- Almost all of the marketers agree that attribution is a problem they’d like solved.
- Only half of the marketers believe that attribution is a problem that can be solved.
So that’s the starting point. From there, the following breakdown of marketers emerges:
Attribution Preachers (20%) -- This group is all in on attribution. They’ve had success in the past with attribution. They’re excited to learn what’s new.
Believers (30%) -- These marketers believe in attribution. They might have had mixed experiences in the past with attribution vendors but their outlook is positive. They’re optimistic that there are attribution solutions that are right for them.
Skeptics (30%) -- These marketers are doubtful of attribution vendors. Some have exclusively had bad vendor experiences in the past and worry about being burned again. Others are content with how they’re currently measuring performance and don’t see the need for anything else. Although they agree that they have the problems that a good attribution product should solve, they don’t think any vendor’s solution is worth the time, money or energy.
Never No-Way’ers (20%) -- These marketers are fully anti-attribution. They’ve worked with vendors in the past. Each vendor failed. These marketers have been burned beyond reproach. To them attribution is dead, cookies are gone, Facebook is a walled garden and they’d be better off watching Netflix than listening to an attribution pitch. These are my favorite conversations.
All of this gets me excited though. Attribution is clearly still a problem in need of solving. Existing solutions haven’t lived up to the hype. To me this is a market that’s ripe for new companies to enter and innovate. At the same time, it’s jarring that 50% of marketers inherently have a skeptical take on attribution. Any company that’s looking to win in attribution will have to overcome this skepticism.
So let's dive into it -- what’s causing the skepticism? Here are the top 10 apprehensions we hear from marketers about attribution.
1) Implementation Hell
This is by far the most common concern we hear. Many marketers have either lived through or heard horror stories of multi-year implementations. They’re not only concerned about how many of their own resources they’ll have to commit but that they’ll have to plead for hard to come by engineering time as well. They’re worried about changing UTMs, getting pixels placed on site and never-ending discrepancies investigations. These are all valid fears that match what we’ve seen firsthand from traditional enterprise attribution vendors.
2) Pricing, Pricing, Pricing
Attribution has always been very expensive. Traditional enterprise contracts start at $200K per year for base functionality. These contracts can quickly get above 7 figures once add-ons and service fees are included -- way too expensive for lots of companies. The other concern is around the pricing model. A lot of traditional enterprise contracts are based on media spend. Marketers hate this. The moment pricing is tied to media, the vendor can no longer be neutral.
3) Low quality data
Everyone wants clean and accurate data. The marketers we talk to have mixed experiences with data quality however. On the one hand, they’ve often been operating with imperfect data in Google Analytics for years. On the other hand, why should they suddenly pay an attribution vendor only to wind up with more inaccurate data? A common ask to address this is to see the attribution solution with their data before having to sign.
4) Can I trust a black box?
Marketers are inherently skeptical. They don’t believe anything unless they can validate it themselves. This makes the idea of using an attribution vendor that just spits out campaign results disconcerting. It’s easier for marketers to just stick to the black box they know. It’s easier to continue to rely on the CPA / ROAS reported out within the platforms they’re used to buying on. After all, the devil you know is better than the devil you don’t.
5) My business is unique. No solution will work for it.
This ties directly in with #4 (“Can I trust a black box?”) -- how can a marketer trust another company to attribute their media when their business is unique? “Will this vendor even be able to fully understand my business? Will this vendor be able to handle the unique aspects of our data?” There is definitely truth to this concern. In fact, I can pretty much guarantee that no 3rd party vendor will ever be able to connect with 100% of the unique aspects of your business.
6) Cookies are dead. Attribution is dead. It’s all about lift / holdout testing.
There are two contradictory trends emerging. It’s confusing and makes it hard for marketers to know where to focus and where to invest their dollars.
Trend #1: There have been endless articles in the last 3 years around the death of 3rd party cookies (actually happening this time), the rise of walled gardens and the resulting death of attribution. Is the time for lift / hold-out testing here?
Trend #2: In the last 3 years both Facebook and Google have launched their own Attribution products. Why are they launching attribution products if holdouts are the answer and attribution is dead? Even more, in what world can I trust Facebook or Google to be my neutral attribution provider?
This concern comes up from marketers that have used an attribution solution in the past. It’s the right question to be asking and a valid concern. These marketers realize that there’s work required to get a sophisticated attribution system up and running. But they’re concerned about what comes afterwards. Is this a never-ending system that needs to be maintained? This ties directly into...
8) How do I get buy-in? Why switch from what’s already working?
“Why even commit to this whole endeavor? We have a system that’s working now. Why maintain something new and why should I try to get buy-in across my company for a new model?” This is a common concern for all marketers. After all, if things have been working to date why rock the boat? There is a more subtle concern here that is not usually spoken of however...the fear of what happens if the new model makes prior media performance look bad. After all, who wants to suddenly go to their boss and say “the marketing we did last year didn’t actually perform.”
9) Different data-sets from different channels (no Facebook view through)
Different channels provide different data sets. Clicks + views are available for display but only clicks are available from Facebook. With TV you can’t get clicks or imps but with OTT you somewhat can? All of this makes it confusing for a marketer to understand how any solution can sit across every channel. I get the confusion -- it’s complicated and there will always be imperfect results.
10) How do I actually use it? How do I prove the ROI?
This is the second most common concern we hear. Marketers are unsure about what they’ll actually do once they have an attribution solution up and running. Even though these marketers know exactly what to do by looking at buying platforms and Google Analytics numbers, the shift to independent attribution feels different. This can feel scary.
How does Rockerbox solve for all of the above? The short answer is we don’t. Although we’ve built products that solve most of the concerns above, there are still areas that Rockerbox doesn’t yet have a perfect answer to. We’ll be writing more about how we address the concerns above but in the meantime, what apprehensions do you have about attribution?