Every B2B founder gets asked the same question at some point: “how long is your sales cycle?” Every B2B founder gives the same answer: whatever the last closed deal was, rounded to a tidy number. The real answer is a distribution, not a single figure, and it is almost always longer than the one people quote.
I am George, founder of Leadpipe. We match anonymous website visits to real people across 5M websites and 280M verified profiles. That gives us a view most analytics stacks do not have: the full tail of research visits a buyer makes before they ever fill out a form or take a meeting. The stretch from the first anonymous visit to the closed-won deal is the actual buyer journey. The visible part, from form fill to close, is the last chapter.
This post is not “here is the average.” It is “here is why the average is wrong, what public research already establishes about B2B buyer behavior, and how to measure your own research window honestly.”
The number you are quoting is the wrong number
When a rep says “our sales cycle is 28 days,” what they mean is “the median time between an opportunity getting created in our CRM and it closing.” That is a real number, but it describes the visible end of the journey, not the journey itself.
The buyer was thinking about you for weeks before that. They visited your homepage on a Tuesday afternoon. They came back Friday and read a comparison post. They spent a day on your pricing page. They ran a search for alternatives, found a competitor, came back to compare. They waited for a board meeting. They asked their CTO to look at it. Then, three weeks in, they filled out a form.
Your CRM logs the form fill as day 1 of a 28-day cycle. Your CRM is wrong. The buyer’s day 1 was 21 days earlier.
This is not a Leadpipe-specific finding. The Gartner B2B Buying Survey has consistently shown that the buying journey is dominated by independent research, with buyers spending around 70% of their evaluation off vendor websites and inside their own buying committees before contacting any vendor. By the time the buyer fills your form, they are usually on the shortlist of two or three vendors and the evaluation is most of the way done.
The implication is that the “cycle” your CRM measures is a fragment of the cycle that actually exists.
Why analytics tools mismeasure this
Three structural reasons your analytics stack cannot give you the right number.
| Mechanism | What breaks | What gets reported |
|---|---|---|
| Cookie expiry / privacy defaults | Identity continuity across sessions | Each visit looks like a new user |
| Form-fill resets | First-touch attribution | Buyer arrives “fresh” on form-fill day |
| Anonymous research | Pre-form sessions invisible to your CRM | Day 1 = day of form fill |
| Multi-device journeys | Buyer on phone, tablet, work laptop | Three “different users,” same person |
The combined effect is that 97% of B2B website visitors leave anonymous, the buyer journey resets to “day 1” the moment a form fills, and the dark-funnel research phase is invisible in your dashboards. Google Analytics is lying about pipeline because GA collapses the buyer journey at form fill and treats the buyer as if they arrived fresh.
This is not a GA-specific problem. HubSpot, Mixpanel, Heap, every session-based analytics tool has the same blind spot. They are built around sessions and events, not around persistent identities. The buyer journey lives in the persistent-identity layer.
What visitor identification reveals
When you run a deterministic identity layer across the same traffic, the picture changes. Sessions that previously looked like four “different users” collapse into one buyer. Anonymous research visits stitch onto the eventual form fill. The pre-form research window becomes measurable.
The rough shape we see across the Leadpipe network is consistent with the Gartner stat: the dark-funnel portion of the journey is typically longer than the visible portion. Buyers spend weeks evaluating before they ever raise their hand. The exact median for your business depends on your category, deal size, and traffic mix, but the pattern is structural.
What we can say with confidence:
- Most eventual buyers visit the site multiple times before filling a form. The return-visit curve study walks through the typical shape.
- The pre-form research window varies sharply by deal size. Sub-$10K ACV deals close fast. Six-figure deals can have research windows measured in months.
- Regulated industries (healthcare IT, financial services, legal) extend research windows because the compliance review itself is a research phase.
What we cannot give you is “the median for your business.” That number lives in your traffic, your CRM, and the link between them.
How to measure it on your own data
Concrete steps. None of this requires a custom data team.
Step 1: Pick a closed-won cohort
Pull every closed-won opportunity from the last two completed quarters. Filter to deals you can actually trace: where the buyer’s account is identifiable and you have at least one logged session before the form fill.
| Filter | Why |
|---|---|
| Closed-won, last two quarters | Recent enough to be representative |
| US B2B (or your geography) | Match-rate dynamics differ by region |
| Pixel active before the deal started | You need pre-form session data |
| Deal size $1K+ ACV | Excludes one-off micro purchases |
For most teams this leaves a cohort of dozens to hundreds of deals, which is enough to see the shape.
Step 2: Stitch the buyer to their research history
For each deal, pull every session attributed to that buyer or their account before the form fill. This is where the identity layer matters: cookie-only stitching breaks across the 30-90 day windows you need.
Three events to capture per deal:
- First observed anonymous visit (day 0).
- First form fill / self-identification (day F).
- Closed-won contract date (day W).
Step 3: Compute three windows
| Window | Calculation | What it tells you |
|---|---|---|
| Pre-form research | F - 0 | The dark-funnel stretch your CRM does not see |
| Form-to-close | W - F | What your CRM calls the “sales cycle” |
| Full journey | W - 0 | The actual buyer journey |
Report all three. The gap between the first column and the second is the size of your blind spot. For most B2B teams the pre-form research window is at least as long as the form-to-close window, often longer.
Step 4: Cohort by deal size
The single biggest driver of research-window length is ACV. Bucket your cohort:
| ACV bucket | Expected pattern |
|---|---|
| <$10K | Short pre-form window; same-day to two-week form fill is common |
| $10-50K | Multi-week pre-form research; one buyer-side stakeholder |
| $50-100K | Multi-month research; multi-stakeholder evaluation |
| $100K+ | Months-to-quarters research; full buying committee |
Enterprise research windows are not 2x SMB. They are usually several times longer, and almost all of the extra time lives in the pre-form-fill phase. Buyers evaluate quietly, build internal consensus, then surface.
Step 5: Count visits, not days
Calendar time is a noisy variable. People go on vacation, projects get deprioritized, quarters close. A cleaner signal is visit count before the first form fill.
Buyer-journey visit-count framework:
1 visit pre-form Hand-raiser, often impulse / referral
2-3 visits pre-form Light evaluation
4-7 visits pre-form Active comparison
8-14 visits pre-form Heavy research, multi-stakeholder
15+ visits pre-form Enterprise / regulated, committee evaluationBuyers who visit many times before filling a form often close at higher rates than single-visit form fills, because by the time they identify themselves they have already done the comparison work. That is the group most undercounted in standard analytics: they look like high-intent new inbounds on the form-fill day and look like nothing in the weeks before it.
This is why the old lead-form model is breaking. See death of the lead form for the broader argument.
Where buyers spend the research window
The pre-form research phase is not random browsing. Across the customers we see, it concentrates on a handful of pages:
| Page type | Why buyers return to it |
|---|---|
| Pricing | Anchoring, comparison, internal stakeholder review |
| Product / feature pages | Fit-checking against requirements |
| Comparison / “alternatives to” | Stress-testing the frontrunner |
| Case studies | Stakeholder convincing |
| Documentation | Technical buyer due diligence |
| Integrations | Stack-fit research |
Pricing-page revisits are the strongest single signal. A buyer who hits your pricing page three or more times across two weeks is in active evaluation. By the time they fill a form, the decision is mostly made. See what to do when someone visits your pricing page for how to engage during the research phase, not after.
Three implications for the operating model
Stop quoting “sales cycle” as your buyer-journey length. If a rep tells you the cycle is 28 days, they mean the 28 days they can see in the CRM. The buyer was thinking about you for weeks before that. If your pipeline forecasting assumes the 28-day view, you are mistiming everything downstream: hiring, quota, capacity planning. Add “days from first observed anonymous visit” to every closed-won deal in your CRM. Put it next to “days from form fill.” The gap between the two numbers is the size of the blind spot.
Start engaging in the research window. The biggest pipeline lever for most B2B sellers is shortening the dark-funnel portion. If buyers research for weeks before identifying themselves, every day you can compress that is compounded revenue. Visitor identification, the deterministic kind, lets you see who is researching and reach out on day 2 instead of waiting for day 30. That is the midbound motion, and it is what replaces the cold outreach playbook.
Measure the full funnel, not the visible funnel. Visit-count and pre-form research window are leading indicators that will move before your sales-cycle number does. Watch them. Report them. The teams that win the next two years will be the ones with the cleanest view of the dark-funnel research phase.
What the journey shape implies for your stack
If the dark-funnel research window is real and material, your stack needs four things it probably does not have today:
- An identity layer. Visitor identification turns anonymous research into a named buyer journey. The deterministic kind, not IP reverse-lookup, is the difference between knowing the company and knowing the person.
- A 90-day measurement window. Anything shorter under-measures enterprise journeys. Anything longer is fine but lags.
- An off-site intent layer. A meaningful chunk of the research window happens off your site. Tools that read intent across the wider web (we built Orbit for this) surface buyers before their first site visit.
- A workflow that engages mid-journey. Knowing a buyer is researching is useless if your motion is “wait for the form.” The whole point of seeing the dark funnel is engaging inside it.
Without those four, you are still operating off the visible 30-day window, while the buyer is in week three of a three-month evaluation.
Limitations of any measurement
Worth flagging before you publish your own number internally:
- Survivorship bias. When you measure closed-won deals, you are measuring the deals that worked. The research window for lost deals is likely longer, because those buyers shopped longer and picked someone else.
- Pixel-on date. You can only see visits after your pixel is deployed. If a buyer visited before that, the visit does not count. This biases your numbers shorter than reality.
- Identity coverage. Deterministic matching on US B2B traffic resolves a meaningful share of visitors, not all of them. Buyers in EU/UK or on heavy-VPN traffic match at lower rates and look shorter than they are.
- Multi-device journeys. A buyer on three devices is one buyer, but stitching them together depends on how good your identity layer is.
The point of measuring it is not to produce a benchmark you can compare to a peer’s number. It is to produce a number you trust, on data you own, that tells you when your buyers actually start thinking about you, not when they finally tell you they did.
Leadpipe identifies 30-40%+ of your US B2B visitors with full contact data on the Pro plan at $147/mo. No credit card to start the 500-lead trial. Start identifying visitors →
