Attribution tools show up in nearly every B2B marketing tech stack. The goal is straightforward: understand which channels are driving pipeline so you can double down on what works and cut what doesn't.

However, no attribution tool on the market is able to accurately track increasingly complex B2B buyers’ journeys, let alone dark social. You may be missing the big picture if you’re basing all your sales and marketing decisions on attribution.

Trinity Nguyen, our VP of Marketing, shared her take on why standard attribution models keep failing revenue teams and what we do at UserGems instead.

UserGems is the AI command center for outbound and ABM. It combines Data Agents, Intelligence Agents, and the AI Chrome Extension to help revenue teams identify who to target, understand what to say, and act at the right moment to generate pipeline and protect revenue. Book a demo to see how UserGems accelerates pipeline and revenue, backed by our money-back guarantee.

Why attribution doesn’t always work

As it becomes critical for marketers to defend their marketing spend, decision-makers rely more on attribution tools to provide the data they need to decide how to allocate capital and show founders and CFOs the return on investment (ROI) of marketing campaigns.

Whether you use a single-touch, first-click, or last-click attribution model, here are some reasons they typically fall short.

1. Increasingly complex sales cycles

With an average of eight touchpoints to get a B2B customer on a discovery call and 24% longer sales cycles, traditional marketing attribution models aren’t able to accurately track a B2B buyer’s journey.

“It's very, very rare that there's one single touchpoint that gives you that deal,” explains Trinity, VP of Marketing at UserGems. “So not the first touchpoint, not the last touchpoint. Some kind of multi-touch is a little bit better, but it's not completely the truth.”

She continues, “For multi-touch attribution models, you also need to have the right tools and set up in place to accurately capture all these touches throughout the cycle. And even then, no tools today (and probably not in the future) can capture the dark social. For example, someone who hears about your product while hanging out with friends or taking out the trash.”

Google Analytics can tell you the last click was a Google ad. It can't tell you the customer had already decided to buy after a conversation with a peer.

2. Unpredictable buyer behavior

A key reason for using different attribution models is to make sure teams get credit for the work they’re doing. And while that sounds like a great idea, the truth is that you can’t control how people find their way to you.

Here’s an example scenario Trinity shared:

“Someone sees our content on LinkedIn, maybe through a customer shout-out. But then, instead of searching for UserGems or clicking on any of our links on social media, they just open a new tab and type ‘usergems.com’ directly. So we have a lot of direct traffic.

“Same thing when sales reach out to them like sales dev reach out to them in the email,” Trinity continues. “We know they opened the email because we can see they opened it. But instead of replying to the rep, they open a tab, type in ‘usergems.com’ directly, go through a website, and then submit a demo request.”

She adds, “I know it firsthand because I was that rep at UserGems. I'm thinking, ‘Hey, I can see you opened my email right now. Why don't you just reply to me?’”

A linear attribution model would report this as direct traffic or demo requests because attribution tools can’t track the other marketing touchpoints or offline channels that influenced the deal.

3. Bigger buying committees

The B2B sales process now involves more people than ever before. The average buying group has grown from 7-9 people to 9-12 people, even in the mid-market.

This means marketers and sales professionals can no longer focus on tracking and nurturing just one person.

“Attribution is hard with a single buyer, but the challenge grows when multiple stakeholders in a buying group each shape the decision at different points in the cycle,” says Trinity.

She explains, “From first principle thinking, and when we consider buying a tool, can we map out our journey? We know it’s not linear. It's not easy to attribute because of this one thing that I bought the product.”

Yet, attribution is essential when it comes to effectively allocating capital to things like ad spend and lead generation campaigns. It provides the metrics needed for data-driven marketing tactics and guides the overall strategy.

So how do you work around these limitations to produce accurate attribution reports?

How we handle attribution at UserGems

At UserGems, we've developed practical workarounds that provide a more accurate picture of attribution.

Instead of using a one-size-fits-all attribution model, Trinity recommends a custom attribution model based on your GTM strategy and ICP’s buying behavior.

Here are a few systems we have in place to get a more accurate picture of where our customers come from.

1. Self-reported attribution

According to Trinity, self-reported attribution has been the most accurate way to track how our customers and prospects hear about UserGems.

We added an optional free text field to our demo request form to ask customers how they heard about UserGems, and to Trinity’s surprise, people actually filled it out.

 “I’m a very anti-long-form person, so I doubted that anyone would fill [it out],” she explains.

“We sell to sales and marketers so they understand the challenge. For example, some customers say they heard about us from a friend, while others say they found us through a Google search and shared the exact keyword they used. They give you the whole fingerprint, the whole map. It's incredible.”

Keeping it optional and open encourages buyers to share their journey honestly, whether it’s through organic search or customer interactions.

As Trinity puts it, “Give your buyer some credit. Some people do want to tell you.”

2. Analyzing closed-won account touchpoints

Another way Trinity attributes revenue is by manually reviewing our closed deals in the past quarter to capture the touchpoints of each customer’s buying journey.

“I look at all the closed-won in the past quarter and identify all of the touchpoints that we captured in Salesforce and HubSpot CR for the contacts we talked to in those opportunities,” says Trinity.

“It's a manual process,” she explains. “I export to a Google Sheet to read, and then start attributing it kind of like, OK, so at some point, this webinar influenced this at some point. I'm not saying that it generated the pipeline or the won accounts and things like that.

“If an account-based marketing program was the one that got the customer in, then both the account development team (ADR),  sales development team, and the campaign team get 1 point in terms of credit.”

Trinity notes:

Deciding who gets credits for a customer can be a source of division for revenue teams, so attribution needs to be efficient and fair.

Here are some methods we apply at UserGems for a more accurate attribution report:

• Giving equal points to teams that influenced a deal

“In many cases, two teams get the same amount of credit for the same deal,” says Trinity. “For example, if the sales development team reaches out to one contact at an account, and then a different person at that account creates an opportunity within three months, both teams get the same points.”

• Using self-reporting and UTM parameters to find touchpoints

Instead of defaulting to first- or last-touch, we combine self-reported data and UTM tracking to map every touchpoint we can across an account.

“When we get a customer, I would look back into the context of that opportunity and trace back through Salesforce and HubSpot,” says Trinity.

“We try to capture and see if, at some point, they came in through a LinkedIn post. So we UTM a lot just to trace where they came from and supplement with any self-reported attribution by anyone in that opportunity.”

She adds, “If every single team touches it, from SDR to campaign to sales and content, they all get a point for it. I'm basically double, triple, quadruple counting it. But to me, that helped assist this deal forward at some point.”

As a result, we can recognize the contributions of all the teams that helped close a deal.

How do you balance scalability with manual attribution?

For larger organizations or marketing teams, you can use automation to scale the attribution model we use at UserGems.

Trinity recommends two ways to scale:

• Contact data: Tools like UserGems automatically enrich and update contact records in your CRM (title, company, seniority, department) so your attribution data reflects who actually moved through the buying process, not stale records that skew your analysis.
• Free text: For free text responses, such as ‘I heard about you from X webinar,’ you can use a tool like OpenAI's GPT to summarize user responses for you. This is a fast and straightforward way to get insights from a large amount of free text data.

Here is an example of how to use GPT to summarize free text attribution data:

• Export the free text data from your CRM
• Put the data into a text file
• Open the GPT website and create an account if you don't already have one
• Paste the text file containing the customer or prospect self-reported responses into the GPT prompt box
• Ask GPT to summarize the text. For example, you could ask, ‘Can you summarize this? or “How often do people say xyz?” Test different prompt phrasings to get more specific outputs.
• GPT will generate a summary of the text

Scan for recurring themes: what source comes up most often, which content pieces get mentioned, and whether a specific channel shows up in self-reports but not your UTM data.

Those gaps are where your attribution model is leaking credit.

Marketing attribution should fit your buyer’s journey

Attribution involves inherent limitations and uncertainty. No model will give you a clean read on every touchpoint that moved a deal forward.

A custom attribution model based on your buyer behavior and GTM strategy gives you a more accurate view of your customer journey.

Through self-reported attribution, analyzing your closed-won accounts for touchpoints, and focusing on attribution as a tool to measure efficiency, you gain a clearer picture of what drives conversions and contributes to your marketing ROI.

UserGems is the AI command center for outbound and ABM. It combines Data Agents, Intelligence Agents, and the AI Chrome Extension to help revenue teams identify who to target, determine what to say, and act at exactly the right moment. Book a demo to see how UserGems accelerates pipeline and drives revenue, backed by our money-back guarantee, or learn more here.

UserGems helps revenue teams find and act on in-market buyers using accurate contact data, AI-scored signals, and AI agents that run coordinated outreach inside your existing stack. UserGems gives revenue teams the data, intelligence, and action layer they need to run high-performing outbound and ABM programs, identifying the right accounts, surfacing the warmest contacts, and delivering prioritized signals directly into existing workflows.

Companies like Mimecast, Greenhouse, and Medallia use UserGems to build pipeline from high-fit, in-market buyers without adding headcount.