How to Track What's Actually Driving Signups: 2026 Guide

TL;DR

Every ad platform claims credit for your signups, but the numbers never add up. Tracking what’s actually driving signups requires layering multiple methods: UTM parameters for click-level data, self-reported attribution for dark funnel signals, product analytics for post-signup behavior, and incrementality testing to prove causation. No single tool will give you the full picture, and that’s okay. Start simple, layer up as you grow, and accept that honest reporting sometimes means “unattributed.”

---

You open Google Analytics. It says organic search is your top channel. You open Meta Ads Manager. It says Facebook deserves credit for 40% of your signups. You open your CRM. Half your leads are tagged “direct.” You check LinkedIn campaign metrics. They claim a bunch of conversions too.

Add it all up, and you’ve apparently generated three times more signups than you actually have.

This is the reality for most startup founders trying to figure out what’s working. Every platform grades its own homework. Dashboards show clicks and impressions, but they don’t tell you what caused someone to sign up. The word “actually” in “how to track what’s actually driving signups” is doing heavy lifting, because most tracking tells you what happened, not why it happened.

This guide defines every key term, method, and tool involved in connecting marketing activity to real signups. It’s organized from foundational concepts to advanced techniques, so you can start wherever matches your stage and build from there.

If you’re running multi-channel campaigns across paid, organic, email, and outbound, this is the reference to bookmark.

---

Marketing Attribution: The Core Concept

Marketing attribution is the process of connecting your marketing channels, campaigns, and touchpoints to actual business outcomes: leads, signups, pipeline, and revenue. For SaaS companies, this typically means tracing how a prospect moves from their first interaction to MQL, trial signup, paid conversion, and sometimes expansion revenue.

Why does this matter so much for startups? Because you’re burning runway. Every dollar spent on a channel that isn’t producing signups is a dollar that could have gone somewhere better. Attribution gives you the data to make those decisions.

But here’s what most attribution content won’t tell you upfront: attribution models estimate credit. They don’t prove causation. Some buyer influences, like peer recommendations, private Slack communities, podcast mentions, and boardroom conversations, will never appear cleanly in any analytics tool. The goal isn’t perfect credit assignment. The goal is making better decisions about where to spend your next dollar.

One common failure mode is operational, not technical. Inconsistent CRM data, unclear stage definitions, and incomplete activity tracking break attribution before any model gets a chance to work. If your team isn’t disciplined about tagging and data hygiene, the fanciest attribution software in the world will produce garbage.

Understanding which lead generation strategies actually fill your pipeline starts with getting this foundation right.

---

Attribution Models Explained

An attribution model is a set of rules for distributing credit across touchpoints. Choosing the right one depends on your sales cycle, deal size, data maturity, and whether your motion is product-led, sales-led, or hybrid.

First-Touch Attribution

First-touch attribution gives 100% of the credit to the initial interaction that brought a prospect to you. If someone first clicked a Google ad, then later attended a webinar, then signed up after reading a blog post, the Google ad gets all the credit.

Best for: Understanding which channels generate awareness and fill the top of the funnel. Early-stage startups with simple funnels should start here.

Blind spot: It ignores everything that happened between discovery and conversion. That webinar might have been the real tipping point, but first-touch will never tell you.

Last-Touch Attribution

Last-touch gives 100% of the credit to the final interaction before conversion. In the example above, the blog post gets all the credit.

Best for: Understanding which programs convert existing leads into signups or opportunities. Useful for sales-led teams focused on bottom-of-funnel optimization.

Blind spot: It systematically over-credits channels that happen to sit at the end of the funnel (retargeting, branded search, email nurture) while ignoring everything that built awareness.

Multi-Touch Attribution

Multi-touch attribution distributes credit across multiple touchpoints. Common variations include:

• Linear: Equal credit to every touchpoint
• U-shaped: Heavy credit to first and last touch, smaller shares to middle interactions
• W-shaped: Heavy credit to first touch, lead creation, and opportunity creation
• Time-decay: More credit to touchpoints closer to conversion

For most SaaS companies with 3 to 6 month sales cycles, multi-touch is the right starting point. Research from House of Martech found that a typical SaaS sale involves roughly 266 touchpoints, and that switching to multi-touch from single-touch models can eliminate more than $50K per year in wasted spend by revealing high-value channels like webinars and ROI calculators that simpler models miss entirely.

Data-Driven Attribution

Data-driven attribution goes further by learning from actual conversion paths rather than following preset rules. It analyzes both converting and non-converting journeys to estimate how much each touchpoint truly changes conversion odds. Google Analytics 4 offers a version of this, though its quality depends on having enough conversion volume to train the model.

Best for: Growth-stage companies with sufficient data volume and multiple active channels.

Blind spot: Still relies on trackable clicks and sessions, so it can’t see the dark funnel.

Quick Comparison

Model	Best For	Primary Blind Spot
First-touch	Top-of-funnel channel discovery	Ignores nurture and conversion touchpoints
Last-touch	Bottom-of-funnel optimization	Over-credits closing channels
Multi-touch	Full-funnel view of complex journeys	Requires clean data across all touchpoints
Data-driven	High-volume, multi-channel programs	Needs significant conversion data; can’t see dark funnel

If you’re running ad experiments across platforms, understanding which model you’re using (and its limitations) determines whether you’ll read the results correctly.

---

UTM Parameters: Your Tracking Foundation

UTM parameters are tags appended to URLs that tell your analytics tool exactly where a click came from. They’re the most basic, most important building block for tracking what’s driving signups.

The five standard parameters are:

1. utm_source: Where the traffic comes from (e.g., google, linkedin, newsletter)
2. utm_medium: The marketing medium (e.g., cpc, email, social)
3. utm_campaign: The specific campaign name (e.g., spring_launch_2025)
4. utm_term: The keyword, if applicable (e.g., saas_attribution_tool)
5. utm_content: Differentiates similar content or links within the same campaign (e.g., hero_cta vs sidebar_cta)

According to HubSpot, 87% of B2B marketers rely on UTM tracking for attribution accuracy. That’s encouraging, but the stat hides a problem: most of those marketers are using UTMs inconsistently.

Common Mistakes

The biggest UTM failure isn’t forgetting to use them. It’s inconsistent naming conventions. When one team member tags a source as “LinkedIn” and another uses “linkedin” and a third writes “li,” your analytics tool treats these as three separate sources. Create a shared naming document and enforce it.

Another frequent problem: not persisting UTM data across sessions. A best practice is to capture UTM parameters when visitors first land on your site and store them in cookies or local storage so they survive if the visitor leaves and returns later.

Why UTMs Aren’t Enough

UTM tracking tells you where clicks came from. It does not tell you which clicks turned into customers or generated revenue. You can see that your LinkedIn campaign drove 1,000 clicks, but without connecting that data to conversions and actual subscriptions, you can’t calculate true ROI.

If you suddenly see a spike in “direct” traffic, it often means tracking broke somewhere and you’re losing attribution data. That’s a red flag worth investigating immediately.

For outbound email campaigns, UTM tagging every link is essential, otherwise you’re flying blind on which sequences and messages produce signups versus just opens.

---

Event-Based Tracking and Product Analytics

Marketing analytics tells you what happens before someone reaches your site. Product analytics tells you what happens after. If you’re only looking at the marketing side, you’re tracking half the story.

Event-based tracking captures specific user actions inside your product: signup completed, workspace created, first teammate invited, first task finished, plan upgraded. These events form a funnel that reveals where users drop off and which acquisition channels produce users who actually activate.

A practical funnel might look like this:

• Step 1: sign_up (account created)
• Step 2: workspace_created
• Step 3: first_invite_sent
• Step 4: first_task_created
• Step 5: upgraded_to_paid

Common tools for this include GA4, Mixpanel, Amplitude, and PostHog. GA4 is free and built for event-based tracking, making it the default starting point for lean teams. Mixpanel and Amplitude offer deeper product analytics with cohort analysis and behavioral segmentation. PostHog is open-source and privacy-friendly.

Practitioners on StackShare describe exactly this need. One CPO wrote: “We need to track which Google ads are leading to signups, and which signups are performing various actions on the site. We then need to visualize this data in various ways.”

That’s the whole point. Tracking what’s actually driving signups means connecting the ad click to the signup to the activation to the paid conversion. Marketing analytics alone can’t do this. You need both layers working together.

To learn how to measure marketing experiments without a dedicated analytics team, that guide covers the practical setup.

---

The Dark Funnel: What You Can’t Track

The dark funnel is everything that influences a buying decision but never produces a trackable click. Word of mouth. A friend’s recommendation in a DM. A podcast mention someone heard while driving. A screenshot shared in a private Slack community. A conversation at a conference. An answer from ChatGPT.

This isn’t a small gap. Gartner’s 2025 research estimates that 70% to 80% of the B2B buying journey now happens in these unmeasurable spaces. That number will keep growing.

As one marketing practitioner put it on digitalosmos.com: “Buyers don’t click anymore, they screenshot and DM. They don’t fill out forms before they’ve made a decision, they self-educate for months and only convert when they’re ready to sign.”

Why “Unattributed” Isn’t Broken

If your reporting shows that 40% of your pipeline is unattributed, your reporting isn’t broken. It’s describing the world honestly. Most content treats unattributed pipeline as a problem to fix. The more accurate framing: it’s the most honest reading of a complex buying journey.

The budget implication is real. Channels you can measure get over-invested in because they’re measurable, not because they’re the most impactful. The most effective B2B teams now deliberately allocate 30% to 40% of their budget to channels with weak attribution but strong known influence: communities, podcasts, events, founder brand building.

AI Search: The Newest Blind Spot

This is the emerging challenge almost nobody talks about yet. When someone asks ChatGPT or Perplexity “what’s a good tool for X” and gets a recommendation, they often visit the website directly. Your analytics records that as “direct traffic.” No referrer. No UTM. No attribution.

Most AI-assisted buying journeys never produce a trackable click, which means your fastest-growing discovery channel might be completely invisible in your dashboards. This makes the next method especially important.

Understanding high-intent account signals beyond traditional attribution data becomes critical when so much of the funnel is dark.

---

Self-Reported Attribution: “How Did You Hear About Us?”

Self-reported attribution is exactly what it sounds like: you ask the buyer where they came from instead of relying on analytics to tell you. It’s the “How did you hear about us?” field on your signup form, post-purchase survey, or onboarding flow.

This method has become far more important with the rise of the dark funnel and AI search. For channels like podcasts, word of mouth, community mentions, and AI recommendations, self-reported attribution is often the only reliable signal you have. Click-based tools undercount these channels by 5 to 10 times.

When to Ask

The best moment is right after the user completes a meaningful action. Immediately after onboarding, after starting a free trial, or after their first purchase. Asking on a landing page before signup creates friction and produces lower-quality answers.

How to Word It

Ask “How did you FIRST hear about us?” rather than “What brought you here today?” The first-heard wording pulls the answer toward origin, the actual discovery channel, rather than the last touch. Keep it a dropdown with an “Other” free-text option. Include specific options for channels your analytics can’t see: podcast, friend/colleague, community, AI assistant.

Known Biases

Self-reported data reflects perceived influence, not causal attribution. Users tend to pick the most familiar option. If someone heard about you on a podcast, then saw a LinkedIn ad, then Googled your name, they might write “Google” because that’s the most recent and recognizable touchpoint. This bias matters when you’re making budget decisions.

The right approach: use self-reported attribution as a cross-check against your analytics data, not as the sole source of truth. One practitioner on Substack put it well: basing your strategy solely on survey results is as damaging as focusing just on last-touch attribution.

---

Server-Side Tracking

Server-side tracking sends conversion data from your server directly to ad platforms and analytics tools, rather than relying on browser-based pixels and cookies.

This matters because privacy changes have crippled traditional tracking. After iOS privacy updates, only about 46% of UK iOS users consent to app tracking. Cookie deprecation, ad blockers, and browser restrictions mean pixel-based tracking misses a growing percentage of conversions.

With server-side tracking, when someone signs up on your site, your server sends that event directly to Google, Meta, or whatever platforms you’re running. This captures conversion data that client-side pixels miss and improves ad platform optimization because the algorithms receive more complete data.

For startups connecting their stack, understanding which integrations matter between your analytics, CRM, and ad platforms is a practical first step.

---

Incrementality Testing

Incrementality testing is the gold standard for proving that a channel or campaign actually caused conversions rather than just touching people who would have converted anyway. It’s the difference between correlation and causation.

Why It Matters

Traditional attribution methods credit campaigns for conversions that would have happened regardless. Your branded search campaign? It’s probably capturing demand that organic search would have captured for free. Your retargeting campaign that looks amazing? Incrementality testing often reveals it’s reaching people who already decided to buy.

Most marketing platforms overstate their impact. Facebook, Google Ads, and LinkedIn all report conversions that include people who would have converted without seeing the ad. Privacy changes make this problem worse because less data means more guesswork in platform algorithms.

How It Works

The basic approach: split your audience into a treatment group (sees the campaign) and a control group (doesn’t see it). Compare conversion rates. The formula is straightforward:

Incrementality = (Treatment Group Results - Control Group Results) / Control Group Results

Common methods include geo experiments (run ads in some cities but not others) and holdout groups (suppress ads to a random percentage of your target audience). According to Forrester, incrementality testing can boost marketing ROI by 30%, primarily by cutting spend on channels that weren’t actually driving results.

Honest Assessment

Incrementality testing requires volume, budget, and patience. You need enough conversions in both groups to reach statistical significance. This makes it impractical for pre-revenue startups or companies with low signup volumes. It’s a growth-stage tool, not a seed-stage tool. But once you have the volume, it’s the most reliable way to know if your marketing spend is actually producing incremental signups.

---

The Layered Attribution Stack

No single method for tracking what’s actually driving signups works alone. UTMs miss the dark funnel. Self-reported attribution has recall bias. Multi-touch models need clean data. Incrementality testing needs volume.

The emerging consensus among practitioners is a hybrid approach combining multiple layers. One Substack practitioner summarized it: “The answer is a hybrid stack combining deterministic multi-touch attribution, self-reported attribution, signal correlation, and incrementality testing.”

Another added the uncomfortable truth: “If you genuinely want to go all ‘dark funnel’ on this thing, you have to be willing and available to talk to a lot of people. Sourcing ongoing qualitative data is what proves that approach out.”

What to Use at Each Stage

Pre-seed / Early stage:

• UTM parameters on everything
• “How did you FIRST hear about us?” field on signup
• GA4 with basic event tracking
• A spreadsheet connecting signup source to activation

Seed / Series A:

• All of the above, plus multi-touch attribution in your CRM
• Product analytics tool (Mixpanel, Amplitude, or PostHog)
• Connect marketing data to subscription revenue (Stripe integration)
• Monthly review of self-reported vs. analytics-reported attribution

Growth stage:

• Add incrementality testing on your top 2-3 channels
• Data-driven attribution models
• Server-side tracking for ad platforms
• Quarterly deep-dives comparing all attribution layers

A critical insight from community discussions: most startups over-buy attribution software before anyone has proven they’ll actually maintain it. If you’re already on HubSpot CRM, use its native attribution reporting first. Only add tools where the existing stack genuinely breaks down.

AgentWeb’s methodology starts with a GTM diagnostic that maps your channels, bottlenecks, and tracking gaps before execution begins, which is exactly the kind of audit that should precede any attribution setup.

---

Connecting Signups to Revenue

Tracking signup volume alone is a vanity metric. The real question: which discovery sources produce paying customers with the highest lifetime value?

A channel that drives 500 signups but zero paid conversions is worse than a channel that drives 50 signups where 20 become paying customers. Yet most attribution content stops at “which channel drove the most signups” and never connects to revenue quality.

To close this gap:

1. Connect marketing data to subscription data. Link your UTM and attribution data to Stripe, Chargebee, or whatever handles billing. This lets you calculate true CAC and LTV by channel.
2. Track activation, not just signup. Which channels produce users who complete onboarding, invite teammates, and actually use the product? A signup that never activates has zero value.
3. Review cohorts by source. Compare 90-day retention rates for users from different channels. You’ll often find that organic and referral users retain at 2 to 3 times the rate of paid users.

This is where tracking what’s actually driving signups becomes tracking what’s actually driving revenue. And that’s the metric that matters.

For a deeper look at connecting marketing efforts to SaaS growth metrics and ROI, that guide covers the full framework.

---

Putting It All Together

The honest summary: perfect attribution doesn’t exist. Every method has blind spots. Every tool has biases. The companies that make the best marketing decisions aren’t the ones with the fanciest dashboards. They’re the ones that combine multiple imperfect signals, review them regularly, and stay willing to act on incomplete information.

Start with UTMs and a “How did you hear about us?” question. Add product analytics to see what happens after signup. Grow into multi-touch attribution and incrementality testing as your volume allows. Accept that 30% to 40% of your best marketing may never show up in a report.

And talk to your customers. The qualitative data, actual conversations with people who signed up, will fill gaps that no tool can.

If you want help building the tracking and execution layer across your channels, explore AgentWeb’s pricing to see which plan fits your stage.

---

Frequently Asked Questions

What’s the simplest way to start tracking what’s driving signups?

Tag every marketing link with UTM parameters (source, medium, campaign), set up event-based tracking in GA4 for your signup action, and add a “How did you FIRST hear about us?” dropdown to your signup flow. This three-part setup takes a few hours and covers both trackable and dark funnel channels.

Why do ad platforms report more conversions than I actually get?

Each platform counts conversions independently, and they all take credit for the same signup if a user interacted with multiple channels. Facebook, Google, and LinkedIn all report the full conversion, not a fractional share. This is why summing platform-reported conversions always exceeds your actual total.

How do I track signups that come from AI search tools like ChatGPT or Perplexity?

You mostly can’t through click-based analytics, because these tools often don’t pass referrer data. Users visit your site directly after reading a recommendation, which shows up as “direct traffic.” Self-reported attribution is currently the most reliable way to capture this channel. Include “AI assistant (ChatGPT, Perplexity, etc.)” as an option in your signup survey.

Is “unattributed” pipeline a sign that my tracking is broken?

Not necessarily. With 70% to 80% of B2B buying happening in unmeasurable channels, a significant percentage of unattributed pipeline is normal and honest. Investigate if the percentage suddenly spikes (that might indicate a tracking issue), but don’t treat a steady 30% to 40% unattributed rate as a failure.

When should I invest in incrementality testing?

Once you’re spending enough on a channel to create statistically significant test and control groups, typically at growth stage. For most startups, that means the channel should be generating at least 100+ conversions per month. Before that threshold, focus on UTMs, self-reported attribution, and product analytics.

Which channels does self-reported attribution capture that analytics tools miss?

Podcasts, word of mouth, community and Slack group recommendations, conference conversations, AI search recommendations, and social media posts that people screenshot or share via DM rather than click. Studies show click-based tools undercount these channels by 5 to 10 times.

Should I use first-touch or last-touch attribution?

Neither one alone is sufficient. First-touch helps you understand which channels create awareness. Last-touch shows which programs close deals. Using both together gives a more complete picture, and multi-touch attribution is better still if your data is clean enough to support it.

How do I connect signup data to actual revenue?

Integrate your attribution data (CRM, analytics) with your billing system (Stripe, Chargebee). This lets you calculate LTV and CAC by channel. Then compare 90-day retention cohorts by acquisition source to find which channels produce not just signups but customers who stick around and pay.