With cookies becoming unreliable and stricter privacy regulations taking hold, marketers face urgent questions about the future of attribution.
For B2B marketers, these disruptions mean more than just changes in technology; they represent a profound shift in how success is tracked, measured and communicated. Marketers are under increasing pressure to prove value and demonstrate clear ROI even as familiar methods become unreliable. Budgets are shrinking, while leadership expectations for precision and accountability have never been higher.
We turned to Drew Smith, founder of Attributa and a recognized expert in B2B marketing attribution, to navigate this evolving landscape. Drew understands the pressures marketers face in this environment of rapid change and high accountability.
In this candid Q&A, Drew cuts through the uncertainty with clear, practical advice on critical topics such as navigating cookie changes, optimizing marketing budgets and integrating AI effectively.
Here’s your straight-talk guide to the new rules of attribution.
“Attribution runs primarily on first-party data. The panic over third-party cookies was always bigger in clients’ minds than in reality.”
Navigating the New Reality of Privacy and Identity
Google paused cookie deprecation again. How is this impacting your clients’ attribution strategies?
It gives clients more confidence about the long-term viability of cookie-based tracking. That being said, Google was going to deprecate 3rd-party cookies, but not 1st party cookies. 1st-party cookies were safe. Attribution runs primarily off 1st-party cookies and 1st-party data. So, attribution wouldn’t have been affected much, if at all, by Google deprecating 3rd-party cookies. But, in the eyes of the client, it was always a concern.
Where should B2B marketers prioritize spending now that first-party cookies have more runway?
My biggest concern with the deprecation of third-party cookies was any data related to intent data. Intent data is powered by a lot of third-party information, including cookies. Without those third-party cookies, intent data would have taken a big hit (unless some other technology is used to provide that information).
So, since third-party cookies aren’t going anywhere anytime soon, you can keep spending money on intent data (although, intent data still isn’t my favorite thing—I find it to be muddy at best). In terms of advertising budget, I’m not the expert to be talking about this, haha.
If regulators eliminate UTM parameters, what’s your recommended alternative?
If this starts happening, then the usage of unique landing pages and/or redirects like bit.ly will come back into vogue again. Or, we’ll see some new technology get developed that allows tracking to happen.
Clean rooms get a lot of hype. When do they truly make sense for B2B marketers?
I think it comes down to accurately documenting your reporting requirements. If you don’t have a requirement that necessitates a data clean room, then you don’t need one.
If you only have one requirement that necessitates a data clean room, then you don’t need one. If you have several requirements that require a data clean room, then you need one. You have to do the work upfront to determine what your reporting requirements are and how to achieve them.
Often, it comes down to asking more advanced questions that existing reporting suites (Salesforce, Marketing Automation, Marketing Attribution, etc.) aren’t built to handle. Once you start asking those more advanced questions of the data, you need to start looking at data warehouses, data lakes and more advanced solutions.
Attribution That Earns Trust: Models and Strategies
CFOs always want proof of marketing impact. What attribution methods quickly build their confidence?
This is a tough one. On one hand, CFOs love incrementality lift tests because of the causality, but that comes at the expense of potentially burning pipeline, which CFOs hate. So, it’s a love/hate relationship for CFOs. Multi-touch attribution (MTA) doesn’t require you to burn pipeline, but it’s also less causal than incrementality.
So, again, there’s another love/hate relationship there.
Here’s the thing: there’s no perfect solution. Media-mix modeling (MMM) has its pros and cons as well. Nothing is perfect, and it really just depends on what tradeoffs a CFO is willing to accept. That means that the way to build trust is through transparency and education, not a measurement methodology.
How can incrementality testing and multi-touch attribution (MTA) work together effectively?
Personally, I view these as complementary. As mentioned, neither methodology is perfect.
I think when organizations use both, they have a fuller picture of what’s working and what’s not working. Incrementality works great for the channels that don’t result in a direct click/touchpoint, such as display and brand marketing. That’s a blind spot for MTA. On the flip side, for things like webinars, that do result in direct touchpoints, the fact that you don’t have to burn pipeline to track/measure them with MTA works really well.
“There’s no perfect solution. The only way to build trust in attribution is through transparency and education—not just methodology.”
When should media-mix modeling (MMM) be added to MTA rather than replacing it?
MMM is predictive in nature, while MTA is not. You can build predictive models with MTA, but it isn’t predictive in and of itself. If you’re trying to get predictive, you need MMM. But, MMM doesn’t get down to campaign level and keyword level tracking, which MTA does, so you can’t just get rid of MTA. You still need that granular level data to make sure that you know exactly which things are working and which things are not working.
Is it possible to accurately forecast ROI before starting an attribution overhaul?
Honestly, you can’t. I tell clients this all the time. The thing is… I can’t predict the future. And, if you want me to try, that’ll cost you about $30k, and I can’t guarantee the results. So, don’t bother.
What’s the biggest warning sign your attribution model urgently needs deduplication?
When we see MQLs from Marketing that never advance to Opportunities. That’s a big red flag. When we see MTA touchpoints (TP) in the wrong position, that’s a big red flag. An example of that latter is there. We recommend that clients create an MTA TP when Sales creates a record. That TP would occupy the first touch and lead creation position. When I see a “Sales Created” TP that occupies a position other than first touch or lead creation, that’s a massive red flag.
AI & Automation: Rethinking Attribution
AI promises automated attribution. What changes should marketers expect soon?
This is already in place at several MTA software platforms. I expect all of them to have this within the next 18 months, or risk being left behind.
What prevents AI-driven attribution from overshadowing human judgment?
The biggest one right now is security. I know many organizations that are unable to utilize AI due to significant legal and security concerns regarding the data fed into an AI platform.
Outside of the legal concern, the biggest thing that is going to keep AI insights from overruling human judgment is context.
There are things that an AI just can’t know. For example, let’s say that we’ve gone to the same tradeshow for 10 years. Suddenly, this year, the attribution from that tradeshow craters. The AI would say to stop going to that tradeshow. However, the human would know that we changed our approach at the tradeshow this year. And, that the solution isn’t to stop going to that tradeshow, but rather to revert back to our previous approach.
Have AI-generated touch weights revealed anything unexpected compared to manual approaches?
I can’t say that I’ve seen any surprises at all. Things change, but I haven’t seen anything change in a way that was super expected or that drastically changed the numbers or the analysis.
Which popular attribution metrics today might seem irrelevant or outdated by 2030?
I think any kind of first-touch attribution is going to be completely unreliable by 2030. For two reasons.
One, the rise of AI-based search. People will search for things on ChatGPT or some other AI platform. That AI platform will recommend a couple of companies to look for, and then the person will go to those companies directly.
Two, the rise of AI-based data. By 2030, we’ll probably be able to build an ICP using a ChatGPT-like platform that automatically grabs names, titles, and contact info and populates it right into your CRM for you, thus making lead sourcing completely obsolete.
“The hardest shift is getting leadership to stop chasing MQLs and focus on revenue.”
Better Attribution in Practice: Tools, Tags and Team Habits
GA4 tags can quickly get messy. What’s your top recommendation for cleaner UTM naming?
Consistency. I only have three rules.
- 1) Be consistent
- 2) Lower case only
- 3) No spaces
But “consistency” is number 1, because if you break rules 2 and 3, if you do it consistently, you’re still okay. So, honestly, just be consistent.
Which popular UTM “best practice” actually makes attribution harder at scale?
Tying your utm_campaign parameter to your Marketo Program Name doesn’t work well for multi-channel campaigns at scale.
When auditing RevOps, what’s the best dashboard to align marketing and sales?
Any dashboard that combines MQLs with Pipeline goal attainment. If marketing and sales do not share the same reality, marketing will be crushing their MQL goals and sales will be missing their pipeline goals. That’s the one.
What’s the most important cultural shift marketers need to make to truly connect marketing efforts to revenue?
Managing up and setting proper expectations with leadership that has gotten addicted to focusing on MQLs.
Adapting Attribution for Modern B2B Journeys
How do you handle attribution for long B2B sales cycles with a lot of off-channel influence?
Honestly, you don’t. You set proper expectations. You acknowledge that a model can’t see the influence that is happening outside of its purview. From there, you either identify a solution that addresses that, or, you move forward with the acknowledgement that you have a blind spot.
How should attribution evolve to fit product-led growth (PLG) models?
The evolution here has to be in how/where we track marketing’s efforts. In a PLG world, much of marketing is happening inside the product itself, as opposed to the exterior. We also have to acknowledge that marketing is going to have an equal amount of impact on “expansion” as it does on “acquisition.” Whereas traditional marketing is primarily focused on “acquisition”. That means that you have to adjust your tracking systems and data set accordingly.
How do you capture the revenue impact of customer success teams in attribution models?
We differentiate their efforts through the use of unique UTMs. We create touchpoints tied to their connected phone calls and email replies. We create touchpoints from their QBRs or other types of regularly occurring business reviews. The goal is also to chart all of the interactions that customer success teams are having.
In addition to that, and this is something that often goes unnoticed, is their impact on churn prevention. You have to create tracking mechanisms that help you understand how their efforts are impacting renewals.
What’s one attribution belief you’ve changed after seeing real client data?
This one will be controversial. I held the belief that marketing helped accelerate deals. The vast majority of real-world client data has proven this wrong.
Sales-only deals tend to move faster than deals that involve marketing. But that’s not marketing’s fault. It’s not that marketing is doing anything wrong. It’s that sales-only deals are the easiest deals an org wins. They include the former customer who moved to a new organization and wants to buy again. They include the relationship-based sale like, “Oh, my sister is your VP of Sales, so I’m buying from you.”
These are the fastest and easiest deals you’ll ever win, and marketing never touches these. These deals skew the data in favor of sales-only deals. Marketing helps win the harder deals.
“Most marketers don’t have a strong foundation in statistics. That’s where we step in—to turn messy data into clear, actionable insight.”
Inside Attributa
What led you to start Attributa?
It was actually a very easy decision. I was essentially building Attributa inside a demand gen agency—the sales, the marketing, the hiring, the strategy, everything—and realized I was building a business for someone else instead of myself. About four days and two phone calls later, I submitted my resignation and started the process of building Attributa.
How has your approach to B2B attribution evolved since launching Attributa?
When I started Attributa, the primary end goal was to get organizations set up and operating on multi-touch attribution (MTA). Now, our philosophy has changed quite a bit, in that MTA is just one tool in a much larger toolbox that includes things like MMM and incrementality. MTA is no longer the end goal, but rather one step in a journey.
In your 20+ years in marketing and data, what’s changed most about how B2B companies use attribution?
Back when I started my journey, we didn’t have much in terms of attribution.
The best organizations had single-touch attribution (lead source specifically.) Nobody really had anything more detailed or complex than that.
The only metrics that most organizations were tracking were things like email opens, link clicks, website heat maps, website conversion rates and those types of metrics.
Since then, the tools (and marketers) have gotten way more complex and are creating far more data than ever before. Nowadays, organizations are much more aware of what their capabilities are than they were back then. And, there’s much more of a focus on maximizing those capabilities.
What’s the biggest challenge companies face with attribution, and how does Attributa help?
I’m going to focus on the biggest one, because it affects all organizations, regardless of your methodology, maturity or whatever. That one is the skills gap in marketing around data, statistics and analysis.
Most organizations do not have a dedicated analyst. Analysis is done by the teams creating the marketing strategies and campaigns, which means their analysis is biased.
Many marketers lack a solid understanding of key statistical concepts, including the difference between mean, median, and mode, and when to use each one. Additionally, they struggle with calculating a win rate, which differs from a conversion rate, and building cohorts. I’m not saying that to pick on or talk down to marketers, it’s just the reality of life. If you aren’t constantly flexing a muscle, it will atrophy.
We help organizations overcome this by guiding them on what to report, defining metrics, building reports, creating analysis frameworks, and occasionally, doing the analysis for them.
We are the expert who knows the statistics and analysis best practices that help organizations create good, clean, actionable data and then turn that data into insights and actions. We’re like the behind-the-scenes superpower for organizations that want to do great things with marketing data, reporting and analysis.
Can you share a client story that shows how Attributa helps companies level up their attribution?
Absolutely. A client that still works with me to this day, that I’ve worked with for going on 9 years now. When I started working with them, they couldn’t even do basic first-touch attribution. Their lead source data was so bad that it was impossible.
Over the course of a couple of years, we fixed their first-touch attribution data, we built them a funnel that allowed them to track conversion rates, velocity and aging at every step in the funnel, we built out their multi-touch attribution system, we created some entry-level predictive modeling and we built out very advanced dashboards for them in Tableau to surface all of this information.
Then, we had meetings every two weeks, with rotating topics, where we reported on and analyzed all of the data. These meetings included the CMO. By the end of this process, the CMO would have more information about what was going on in the organization than the CEO, Head of Sales and the CFO.
We also conducted meetings every two weeks with their SEO/Paid Media agency (Noble Studios) to share our data and insights as well. And these meetings allowed us to proactively identify campaigns that could be optimized before too much money was wasted (even the best swing and miss sometimes). This org went from zero reporting to a very mature reporting and data ecosystem over the course of about two years.
Meet Drew Smith
Drew Smith is the Founder/CEO of Attributa, a consulting firm that specializes in marketing reporting, data, attribution and analytics. With more than 20 years of experience in marketing and sales–10 of which were spent consulting–Drew has helped a large number of organizations navigate the complex and tricky waters of marketing reporting and data. Drew is always on the lookout for new methodologies or practices to help marketers understand what’s working and what’s not working, and employs them with his clients.
