Modules · Meeton Ads
Researchers & evaluators
Most visitors to a B2B site are in one of two modes. Researchers are learning about the problem: blog posts, guides, editorial content. Evaluators are deciding about you: pricing, features, comparisons, security, case studies. They convert at very different rates and they want different things, so a journey built for one of them is wrong for the other. This page is the research behind that split and how to turn it into journeys.
The split follows the classic marketing funnel: awareness content (blog, guides) is where researchers are; consideration and decision content (product, comparisons, pricing, demo) is where evaluators are. A third group, implementers, are existing users in your docs and help centre, and a fourth is everyone else: careers, press, about. "Researcher" and "evaluator" are our shorthand for the first two; "buyer intent" and "funnel stage" are the industry's.
What the research found
DynaMeet looked at engaged visitors (three or more page views) on three customer sites over 60-day windows, grouped each visitor by the kind of content that dominated their visit, and measured who later became a contact. The three sites had different shapes: one blog-heavy, one with general traffic, one built around a single editorial hub.
| Site | Evaluators convert | Researchers convert | Gap | Share of engaged visitors |
|---|---|---|---|---|
| A · blog-heavy | 5.1% | 0.3% | 17× | 30% evaluators · 59% researchers |
| B · general traffic | 16.2% | 2.0% | 8× | 31% evaluators · 12% researchers |
| C · one editorial hub | 17.0% | 0.2% | 85× | 8% evaluators · 81% researchers |
- Evaluators convert after 3 or 4 pages. The median evaluator who became a contact had read three or four pages first. Pricing, then security, then a comparison is a typical path.
- Researchers are shallow and plentiful. They cluster at two pages: land on a post, read one more, leave. On editorial-heavy sites they are 59 to 81% of engaged traffic.
- Depth is not the problem; content is. Conversion rises with depth on all three sites, up to 10 to 19 pages. A visitor who reads ten blog posts is still a researcher; a visitor who reads three product pages is an evaluator. The discriminator is what they read, not how much.
- The transition is the moment. Of the few researchers who did convert, three in four had left the editorial section for product or pricing pages first. Researchers become evaluators; the first product or pricing page is where it happens.
- Interrupting evaluators costs leads. On one site, visitors shown on-site ads converted at 1.6% against 3.9% for a no-ads holdout over 19 days. A second site was inconclusive.
These are directional findings from three anonymised sites and a path heuristic, not a law. The 8 to 85× gap held on every site tested; the exact rates will not match yours. The claim that researchers respond better to a soft ask is a hypothesis the team wrote down to test, not a finding.
What Meeton does and does not detect
Meeton does not label visitors as researchers or evaluators. An earlier version classified each page and tagged visitors by the furthest stage they had reached; it was removed because the hypothesis did not hold: not everyone on a pricing page is evaluating, and some people convert after one page. Intent turned out to be a property of the person, their maturity in your subject, not of the page they are on. What survives is the thing that does carry the signal: the path. A journey is a path condition, pricing and then security, and that ordering is exactly the researcher-to-evaluator distinction. So the modes are how you design journeys, and journeys are how Meeton acts on them.
Two live features touch the same idea without using the words: the intent level on a visitor (High once they view a purchase page after researching or read one in depth; Medium for a research page read in depth or a return visit; Low for a first look at a research page), and the journey audience (Visitors, Contacts, Booked a meeting).
The rule
Interrupt the people who would not convert anyway; get out of the way of the people who will.
| Mode | Converts | Treatment | Never |
|---|---|---|---|
| Researcher | 0.2 to 2% | Capture. Offer more content for an email: a guide, a checklist, a benchmark, "notify me". Then nurture toward evaluation with Library and Email. | Ask for a meeting. The ask is wrong, not the person. |
| Evaluator, early | 5 to 17% | Answer "versus what?" Comparison content, a customer story, a side-by-side of price, security and integrations. | Another blog post, which reads as condescending and wastes the slot. |
| Evaluator, ready | Subtract. Stop interrupting, make the low-friction ask easy: "ask a question" or a 15-minute slot, and notify the owner. | Loud popups on page load. | |
| Implementer · other | — | Suppress. No ad, no ask, on docs, help, careers and press pages. | Anything. |
A ladder, not a ratchet: if the meeting ask is ignored, fall back to content rather than asking again louder.
Turning it into journeys
Start with Analyze traffic on the Ads tab: it draws the paths visitors actually take, and the editorial section and the product section usually separate on sight. Then build one journey per mode. Everything below uses the live editor: pages, placements, moments and audiences and creative actions.
A researcher journey: capture
/blog/* or /guides/*. Step 2: /*, so the journey fires on the second editorial page, after they have shown they read rather than bounced./pricing, /product/*: a visitor who has already touched evaluation content belongs to the next journey.An evaluator journey: subtract, then ask
/pricing. Step 2: /security, /compare/* or /customers/*. The second step is the "versus what?" moment; two evaluation pages in order is the signal./pricing and /security so the owner can step in.The transition: a researcher who crosses over
The most valuable event on the site is a researcher opening their first product or pricing page. Build it as its own path: step 1 /blog/*, step 2 /pricing or /product/*. On step 2 offer comparison content, not a meeting yet: they just arrived in evaluation mode. The journey that asks for the meeting starts one page later.
What to suppress
Put /docs/*, /help/*, /careers/* on every journey's excluded pages, or leave those sections out of every path, and keep the format switches under Settings → Meeton AI → Ads for the rest. Implementers are customers; a popup there costs goodwill and earns nothing.
Reading the results
- Compare journeys on leads per 1,000 reached, not on clicks: a researcher journey will have a lower rate and that is expected. See Results.
- Quick closes on an evaluator journey mean you are interrupting people who were about to convert anyway. Move the placement later on the path or lower its loudness.
Troubleshooting
| Symptom | Likely cause | What to do |
|---|---|---|
| The researcher journey has reach but almost no leads | Expected at 0.2 to 2%; or the ask is a meeting | Check the creative action is content-for-email, and judge it on emails captured and later Library reads, not on bookings. |
| Evaluators close the popup at once | The ask comes too early or too loud | Move the placement to step 2 or 3, switch page-load to time-on-page or scroll, lower How loud. |
| Both journeys fire on the same visitor | Their first pages overlap, or the exclusions are missing | Exclude evaluation pages from the researcher path and vice versa; a visitor walks one journey at a time, entering at its first page. |
| Your site has no clear editorial section | Site B's shape: mostly general traffic | Skip the researcher journey; build the evaluator path and a transition path from the home page. |
| Analyze traffic shows no common path | Hundreds of distinct paths is normal; one site had 951 shapes across 1,904 visitors | Fold sections with /blog/*, slice by device, and build on the two or three thickest ribbons. |
Frequently asked questions
Can Meeton tell me which visitors are evaluators?
Not as a label. The intent level on each visitor and the journey they walked tell you the same thing: a visitor with High intent has viewed a purchase page after researching. Journeys are how you act on the mode without needing the label.
Why not just rule that /pricing means evaluator?
Because it misfires: a researcher glancing at pricing is not evaluating, and some people convert from one page. The ordered path, pricing and then security, is a much stronger signal than any single page, which is what journeys express.
Should researchers get no ads at all?
They should get a different ask: content for an email, never a meeting. Compare its leads per 1,000 reached against your evaluator journeys to see how it does on your site.
Where do implementers go?
To the chat agent, which answers their questions from your knowledge. Keep ads off docs and help pages.
Is this specific to B2B?
The research was on B2B sites with editorial content. The two modes exist wherever a purchase takes research, but the rates and the content types will differ.