5. Attribution
(the transcript is below)
Action Step
Complete this before moving on.
Watch the full overview video above and read through the transcript below. Pay attention to the difference between lead source and channel, and why solid first-touch and MQL-touch tracking provides 95% of the value most teams need.
Part 1: Hook / Open
A CMO walks into a board meeting and gets asked "what's the ROI on paid search versus events?"
The honest answer is... they don't really know.
They might have a sense, they might have a spreadsheet someone put together last quarter, but they don't have the infrastructure that tracks this automatically.
That's a pattern we see a lot — marketing is spending, sales is closing, but nobody can connect the two.
And the project that solves it is what we call Attribution.
The attribution project is the infrastructure that captures where every lead came from, what channel drove them, and whether that channel actually produces pipeline and revenue.
My name is Yasin from LeanScale, and in this video I'm going to break down for you our entire Attribution Playbook... what it is, the core concepts, how it gets built, and what changes once it's in place.
Part 2: What Attribution Is
So when we talk about the attribution project, what we're actually talking about under the hood is building the fields, the workflows, and the automations inside of a CRM and marketing automation platform that automatically capture where every lead came from and what action they took.
On the input side of this project there are marketing channels, lead sources, UTM parameters from ad campaigns, and existing CRM data.
On the output side, once the project is complete, every single lead that comes through gets stamped with exactly where they came from and what they did, and every opportunity gets stamped whether it's inbound or outbound, and there are dashboards that show conversion rates by channel, by campaign, and by source.
Now the beautiful thing is once it's built, the data capture runs in the background.
Every form fill, every campaign click, every event registration gets captured automatically — no one is updating a spreadsheet, no one is guessing.
The way I like to describe it is that attribution is like a flashlight.
Without it, it's a dark room — the overall shape of things is visible but the details aren't clear. Leads are coming in but there's no way to clearly identify which channels are actually driving the ones that close.
The attribution project is the flashlight that lets everyone see where the investment is actually going.
Part 3: Attribution Pro Tips
Now the Playbooks Library below this video goes deep on the full methodology behind the attribution project — but before we get into how it gets built, here are a couple things you really don't want to get wrong when implementing this into a startup.
First — understanding the difference between lead source and channel is key. Almost every company we work with has them combined into one field in the CRM, and that's a problem. Here's the difference between them: a lead source is the action someone takes — demo request, content download, webinar registration — a channel is where they came from — paid search, organic social, direct. If those live in one field, the data is useless because there's no way to answer "which channel drives the most demo requests?" — they have to be separate, always.
Second — multi-touch attribution tools that cost anywhere from $12,000 to $130,000 a year... give about 5% of the value most teams are actually looking for. The other 95% comes from solid first-touch and MQL-touch tracking, which is exactly what the attribution project builds — so build the foundation first, and most teams find that foundation gives them everything they need.
Third — attribution is not a set-it-and-forget-it project — it requires ongoing maintenance, and we'll talk about why in a minute.
Part 4: The Problem in Context
And the data backs up why this project is so key for startups who are scaling.
According to McKinsey, 76% of marketers say they struggle to determine which channels deserve credit for conversions.
Ruler Analytics found that 42% of marketers are still doing attribution manually in spreadsheets. Only 31% say they're confident in their own attribution data.
The real cost: the Digital Marketing Institute found that companies without proper attribution misallocate up to 30% of their marketing budget. So if a company is spending a million a year on marketing, up to $300,000 could be going to the wrong place and there's no way to know.
And it's not just marketing that feels this — sales can't see which lead sources produce the best quality opportunities, finance can't get clean CAC numbers by channel for the board, and leadership is setting targets without understanding where the pipeline is actually going to come from.
Part 5: How It Gets Built
So how does an attribution project actually get implemented?
At LeanScale, for all of our projects, we follow a four-phase approach — Strategy, Engineering, Enablement, and Handoff.
Strategy
In the attribution project, like in most of our projects, strategy is the phase that matters most.
This is where we work with all the stakeholders of this project — the CMO, the head of demand gen, marketing ops, sometimes the sales leader — step back and define the lead source taxonomy, the channel mapping, the UTM standards, and the methodology for determining whether something is inbound versus outbound.
The reason this is critical is that if the taxonomy is wrong, everything built on top of it is wrong too... so nothing gets built inside the systems until the strategy is locked.
Before the project even kicks off, our team scrapes the company's existing channels — the website, the ad platforms, the forms — and assembles a V1 of what the attribution taxonomy needs to look like, based off best practices from having implemented this project for dozens of startups before, combined with the unique data from the team we're working with.
We also lean heavily on AI agents throughout this process — agents trained on our playbooks that know how to scrape that information from a startup's existing channels and ads and put together a V1 strategic handout. We go deeper into how we use AI in the implementation of this project in the full playbooks in the library, so if you're curious you can check that out.
From there, we iterate on that V1 with the stakeholders — CMO, demand gen lead, marketing ops — and at the end of that process, four things are locked in: the lead source taxonomy, the channel and sub-channel mapping, the UTM translation table, and the methodology for determining whether an opportunity is inbound or outbound.
Engineering
We then move on to the next phase, engineering — this is the technical build.
This is where the fields, the workflows, and the automations actually get created inside the systems.
We have a 12-step build sequence that covers everything from creating the attribution fields, to building the channel translator — which is the centralized automation that takes raw UTM data and converts it into clean, reportable channel values — to setting up the stamping logic for first touch, MQL touch, and latest touch, all the way through to opportunity attribution.
So what's changed in the agentic AI era for when we do engineering implementation for this project?
We've been over the last few months building AI agents that connect to each of the major systems in go-to-market. We've built an Attribution Agent that uses JSON files in a database with all of the field specifications, workflow logic, and build steps codified. Using APIs and MCPs, that agent connects directly to the CRM — whether that's Salesforce or HubSpot — and the major marketing automation platforms, and starts building out the data architecture from the strategic output.
What used to be purely a 40 to 50 hour manual engineering effort now gets accelerated because the agents handle the repetitive build work while our human engineers focus on quality assurance, orchestration and the edge cases.
The full details of the AI agents we use and the 12-step build process are broken down inside the playbook library if you're curious.
Enablement
Phase three is enablement, because the system is useless if the team doesn't know how to use it.
For the attribution project:
We train marketing leadership on how to read the dashboards — what the conversion rates mean, how to identify which channels are working, how to make budget decisions from the data.
We train demand gen and any ad agencies on the UTM standards — what parameters to use, why consistency matters, and what breaks if those standards change.
And we train marketing ops on how to maintain the system going forward — how to add new channels to the translator, how to monitor the data quality lists, how to clone workflows when new campaigns launch.
Handoff
From there, phase four is the handoff, and this is where maintenance expectations get set.
Now one thing to understand — attribution is not a one-time project.
It requires five to ten hours a month of ongoing maintenance to keep the data quality where it needs to be.
And here's why that matters: every time a new campaign launches, every time an ad agency partner changes a UTM value, every time a new channel gets added — someone needs to update the system.
If nobody owns that, what happens is new leads start flowing in with UTM values the system doesn't recognize. Those leads get bucketed under "unknown" or "direct" — which means the data starts telling the wrong story.
Over the course of a quarter, that compounds — and suddenly the dashboards show that 30% of leads are "direct" when in reality they came from paid search or LinkedIn, the tracking just wasn't updated.
That's how teams end up back where they started, making gut decisions with dirty data.
So the handoff isn't just about transferring ownership — it's about making sure after we've completed the implementation... someone on the team understands the system deeply enough to keep it accurate.
Part 6: What It Unlocks + Close
So let's bring this all together to where we started. Once the attribution project is in place, what actually changes?
Marketing budget decisions go from gut feel to data — which channels convert, which campaigns are worth scaling, which ones to cut.
Companies with proper attribution see a 15 to 20% improvement in marketing ROI and reduce wasted ad spend by 27% on average.
And the biggest thing that changes after this project is confidence and certainty.
A CMO can walk into a board meeting and answer the ROI question with real numbers.
The demand gen team can optimize based on what's actually working instead of what feels like it's working. And finance gets clean data without anyone manually building reports.
That's what attribution does — it takes the operation from guessing to knowing.
If you want more information about what we've covered here — the concepts, the methodology, the full implementation process — all of it is broken down in detail in our Playbooks Library for you to go through. Our Advisory Overview Playbook covers the problem, approaches, and strategic understanding behind this project. The Methodology Playbook goes deep on every concept we talked about for this project. And the Implementation walks through the step-by-step build process.
And if you're a revenue leader at a fast growing startup who's feeling good about the attribution side after watching this — but you're thinking, okay, now how do I make sure reps are actually targeting the right accounts based on what's working — we have a whole playbook on exactly that called the Market Map. It's broken down the same way in our Playbook Library. And while you're there, you'll see we have playbooks on every major GTM project — from Growth Model and Automated Inbound to Quote to Cash and more. Feel free to check those out next.
Again, this is Yasin from LeanScale, and I'll see you in the next one!
Comment in Slack
Post your answer in your onboarding channel.
What did you learn? What clicked?
Any questions — or will you have more questions once you actually get into the hands-on trainings? Either is fine. Just capture where you are right now.