It’s Time To Fix Your Enterprise Revenue Attribution Model
April 4, 2023
Enterprise marketing is broken.
It’s broken because the way buyers discover products no longer matches up with how we’re trying to reach them. Buyers are everyday people, increasingly younger people like Gen Zers and millennials (although do we still count as young?). They might hear about your product from a friend, come across your brand on a podcast, stumble upon a LinkedIn post, or catch a hot thread in the comments of a virtual event.
Meanwhile, our digital attribution systems are tagging those buyers as being generated by channels like organic search or direct traffic to the website. This causes a disconnect between where we decide to invest to go find more enterprise buyers and what actually moves the needle.
All of this matters—even for PLG companies. While PLG companies initially target end users of their products, the goal is to go from individual user to team to company-wide adoption as quickly as possible. A large share of revenue ultimately comes from a subset of enterprise accounts, even if those accounts landed on a free or low-ACV plan. Mature PLG companies must find a way to create enterprise demand for their products, all while sustaining their product-led marketing efforts.
Nobody has been as vocal about this topic as Chris Walker, the outspoken CEO of Refine Labs. Chris has the experience and data to back up his 🌶️ takes. Refine Labs works with 200+ high growth B2B companies, including PLG companies like Algolia and SEON, to help them strategize, measure, and sell into their most valuable accounts. The shorthand: it’s similar to account-based marketing (ABM), but broader and more all-encompassing than what ABM has become known for.
I caught up with Chris to explore why traditional enterprise marketing no longer works and what modern SaaS companies should do instead. Keep reading for insights into the difference between demand creation and demand capture, the right way to do attribution, and why gated content no longer serves B2B SaaS. (The conversation has been lightly edited for clarity.)
On the big shifts in B2B buying
B2B buyers trust what their peers say more than other sources of information. This was happening before COVID, but it’s been dramatically accelerated by COVID.
Instead of going to Google and searching “what tools should I buy for this?” I’m going to go into a community and I’m going to say it to the 13 CMOs that are in this community with me. Or I’m going to text three of the people that I met through this conference and I’m going to see what they’re doing. Some people just post on LinkedIn, “Hey, I’m looking for a sales engagement tool. Which one should I buy?” and get a hundred comments back from people.
People are still searching for something, they just have a much wider range of places that they can search from. It was typically Google 10 years ago. Now that search can happen on Instagram, TikTok, LinkedIn, in a community, inside of a LinkedIn group, through a text message or a phone call or a webinar. In the chat of a webinar is a place where a lot of products actually get discovered. And people trust these sources more than the results they get in Google. If you look at the gross numbers about SEO and SEM, search isn’t declining, but it is less effective for marketers.
On why marketers don’t capitalize on these shifts because of broken attribution models
Another part is that the activities that are used typically don’t result in a direct response conversion that’s trackable with UTMs or things like that. Channels that are typically used for what we call demand creation—going out and getting net new accounts to be aware of your company and aware of your differentiation before they talk to a sales rep—don’t get prioritized because attribution software is based on trackable digital touch points. It limits the go-to-market team on the available things that they have in their tool belt to deploy to try and get more customers.
We did a year-long study and looked at $21.5 million in closed-won revenue. And we compared what HubSpot attribution software said to if we just asked the customer how they heard about us on the sign-up form before they started a sales conversation. (This is not a knock on HubSpot—all digital attribution platforms have the same limitations when it comes to privacy policies and non-direct response conversions.)
In our data, HubSpot attribution showed that 80% of revenue came from organic search, direct traffic or paid search. And when we asked customers, it was 3%. Attribution software is the primary mechanism that companies use to measure the success of their marketing activities, but it’s not actually picking up what the customer says is making the impact.
When you ask the customer, 41% of revenue came from social media and 53% of revenue came from podcasts. Attribution software had 3% for social media and zero for podcasts. There was also a huge amount of spread where 21% was due to what the customer said was word of mouth, and our data had 3% from attribution software.
The finding here is that attribution software dramatically overweights the impact of organic search, direct traffic and paid search due to digital attribution bias. Demand creation activities, or what we call a lot of dark social activities (social media, podcasts, word of mouth, communities), do not get properly measured by attribution software. Some of these things don’t get measured by attribution software at all.
These channels that are highly effective for customers and what customers are saying are driving them to buy are not being properly measured and therefore not being invested in by companies.
On how attribution bias leads to company <> customer misalignment
This causes sales and marketing misalignment, but the underlying driver is that it causes company <> customer misalignment. It causes company <> customer misalignment because the company can only do certain things that get appropriately measured.
Companies have to do some level of a transactional digital event if they want it to be measured and attributed properly, which then leads to lead generation programs that go to email nurtures, or a lot of times direct sales outreach. Then you have your sales team calling people that didn’t ask to be called and are not ready to buy, creating misalignment between sales and the customer as well as the company and the customer.
The second part is it restricts the company from using the things that customers say are most effective. Social media gets deprioritized due to the lack of measurement, podcasts get deprioritized. Communities, whether you build one yourself or you have evangelists engaging in other ones, get deprioritized. Sales reps or customer success reps are not incentivized to build personal brands on LinkedIn, which drives impact and awareness to the company.
On the way forward for building enterprise pipeline
The first thing is level-setting on what is the point of attribution. Over time, the evolution of digital attribution technologies has led people to believe that they’re going to get the exact right result and then be told exactly what to do. The reality is that when you’re developing a go-to-market strategy, you’re getting insights from a lot of different areas and you have to synthesize that in your brain and make good decisions.
There’s not a perfect roadmap of what to do. It’s figuring out what are the sources of information that I need to make good decisions, why do I need them, are they reliable data sources, and then how do I use that information to adjust my strategy or make decisions?
What we are recommending to companies is what we call a hybrid attribution model, leveraging multiple sources of data to synthesize and make good decisions for the go-to-market overall.
One input to that can be digital attribution software, whether you use a last-touch model, or any model that you want. That’s one input, and what it’s typically giving you is how are you effectively capturing demand? Are you capturing demand through events, is it through Google paid or organic? Is it coming from content syndication or lead gen programs? You can see how the demand is being captured.
Then we need a method to think about how are we actually creating demand? The simplest and easily most implemented is adding a “how did you hear about us?” question inside of your high intent forms where a lot of revenue comes through on your website. In some PLG companies, you might not want to put this on the free trial. You do some A/B tests depending on how much volume you’re doing through that form, but definitely on a contact sales form, this makes total sense.
Use that information of what customers are saying as a way to interpret what’s creating the demand. You’ll hear responses like podcasts, LinkedIn, I’ve been going to your virtual events for the past six months, this person told me about you, I moved from one company to another so I’m bringing you here, and so on. You get a lot of that interesting data that’ll never show up in attribution software.
Then on top of that, you can layer on additional methods based on how advanced you want to go. For instance, you might have a large scale market research survey that goes out to a subset of your buyers on a quarterly or bi-annual basis about where are they getting information, what sources do they trust, where did they hear about your company, what percentage of them know what your company does and can recall that. That can be a great way to understand for people that may not be in a buying cycle right now, do they know you? And if they do know you, where are they hearing about you? Which can be a good driver that’s far away from the sales pipeline and definitely going to be difficult to measure.
On the most important KPIs for your enterprise team
This is totally dependent on your go-to-market strategy. What the marketing team should focus on is what your business most cares about.
But if we simplify this in a PLG motion, the setup that we found works well is to have a growth or PLG team and then an enterprise team. They’re doing two different things.
The enterprise team is focused on how we get enterprise customers on enterprise plans. And we probably have a defined set of accounts to know exactly who they are and then we should score marketing on how much pipeline we are creating through these accounts regardless of how they come through. Are they coming through a free trial, contact sales, expansion of current customers? Any of those different motions I think can be important and then you can decide how to prioritize them.
On the PLG side, I think that the marketing team has to be focused on either activated or paying users, something further away from free trial or sign up only. We have to go a little bit further than that and then we have to have some measure of CAC payback on these types of users. PLG companies might get a user paying only $7 a month, but they’re spending a lot more than that to acquire the customer. I’ve seen a lot of PLG companies take that area of the budget and they just optimize for the total number of signups, and they waste a lot of money on people that never convert.
On whether to gate your content
My thinking on this is no. I know some people will disagree with me on this, but the core reason I think this is because it fundamentally changes your content strategy.
- It changes what mediums you create content in—you’re not just stuck in PDF land.
- It changes the actual type of content that you create because the goal isn’t to create an attractive headline and get a bunch of people to put their email address in for it, the goal is to drive people to want to buy stuff.
- It completely changes your distribution strategy. Instead of putting out gated PDF content, you could take that information and somehow repackage and redistribute it on LinkedIn, podcasts, live events, places like that, which don’t require a signup.
Gated content is all about being able to track things and create leads. Two things that I don’t think are serving B2B companies anymore.
On the impact of ChatGPT on marketing
My position on this is it’s actually the best thing that’s ever happened to content and marketing.
The reason being is that ChatGPT is really good at regurgitating known information, and clearly not strong at this point in creating net new valuable information. What this does is it’s going to separate the difference between whether this content creator is producing non-obvious, valuable, unique insights, or are they regurgitating existing insights?
I believe we’ll see a separation in the spread of content creators where people get highly rewarded for bringing proprietary data in for interpreting what companies struggle with, and then developing a net new framework that companies haven’t used before. The creation of net new information will become much more valuable than it is today.
On the signals that a founder actually ‘gets’ marketing
There’s a variety of these. One of the most obvious ones is, does the CEO or co-founder operate on LinkedIn frequently, or operate on some social network frequently? If they do, that means that they value dark social and they think progressively about how to actually engage with customers. They recognize that there’s a huge impact, even though in their measurement system of attribution, they’re probably not measuring that right now.
Just generally looking and watching what they’re doing is so interesting as well. You can go on LinkedIn and see every ad a company runs. Is it all demand gen ads to download eBooks? Or are they thinking differently and boosting content from their CEO or boosting a video from a customer? What are they doing there is an interesting one.
If you can, it’s really interesting to go through their sales process and see what it’s like.
And then how do they think about attribution and measurement? I’m pretty convinced that if a company’s like, “We need to see everything directly attributed by this tool, or the goal is 5,000 leads,” some of those become really negative signals to me.
We should be looking at the business outcomes, not the leading indicator to business outcomes when we think about marketing leadership.
The TL;DR – Seven insights on enterprise pipeline for PLG products
- PLG teams need to create enterprise demand all while sustaining their bottom-up, product-led marketing.
- B2B buyers trust what their peers say more than other sources of information. They’re searching for and learning about products on social media, communities, and peer networks—not just trackable digital channels.
- Digital attribution tools don’t properly measure newer channels, which leaves marketers blind to what actually creates enterprise demand.
- Set up a hybrid attribution model that gives you visibility into both demand creation as well as demand capture. This can be as simple as adding a “where did you hear about us?” question on your high-intent forms for enterprise buyers.
- Enterprise marketing teams should focus on pipeline creation from a defined set of accounts. This should be independent of how those enterprises come through, whether that’s a free trial, contact sales form, or expansion from current customers.
- PLG marketing teams should focus on either activated or paying users, not just all signups. And they need to look closely at CAC payback on these users.
- Marketing leadership needs to focus on business outcomes, not the leading indicator to business outcomes. Don’t let broken attribution models hold you back from generating enterprise revenue.