62% of software buyers now want to meet a vendor's salesperson only in the evaluation or decision stage, up 17 points in a year, according to G2's 2025 Buyer Behavior Report, a survey of 1,169 B2B software decision makers in April 2025. The research stage, the longest part of the journey, is increasingly one buyers want to finish alone, using review sites, AI search and the vendor's own website.
This page collects the dated numbers on that journey for B2B SaaS: how long it runs, how much buyers want to do without a rep, and which touchpoints decide the shortlist. Each figure names its publisher, its sample and, where the publisher sells something related to the finding, that fact. Our broader B2B Website Statistics 2026 post covers the cross-industry buyer data; this one stays on software and on the website's role.
How long the B2B SaaS buyer journey runs
Journey length depends on where the clock starts. Survey data asks buyers how long their purchase took; attribution data measures from the first tracked touch to closed revenue, and so runs longer.
| Measure | Figure | Source and sample |
|---|---|---|
| Average buying cycle | 10.1 months (11.3 in 2024) | 6sense, nearly 4,000 buyers, 2025 |
| Time spent evaluating each vendor | 2.0 months (2.6 the year before) | 6sense, 2025 |
| Software cycles against physical goods | 2.8 months longer | 6sense, 2025 |
| First marketing touch to closed revenue | 272 days (211 the year before) | Dreamdata, 3.5 million journeys, August 2024 to July 2025 |
| Large companies, over 250 staff | 326 days | Dreamdata, same data |
| MQL to SQL | 92 days (107 the year before) | Dreamdata, same data |
| SQL to closed-won | 52 days (62 the year before) | Dreamdata, same data |
| Share of the journey before the sales pipeline | 81% (70% the year before) | Dreamdata, same data |
Source: 6sense 2025 B2B Buyer Experience Report; Dreamdata LinkedIn Ads Benchmarks Report 2026, as reported by PPC Land
Both sources are vendors. 6sense sells buyer-intent software and Dreamdata sells attribution software, so each has an interest in showing a long, mostly invisible journey. Their data still agrees on direction: the total journey is not getting shorter, but the part a salesperson sees is. In Dreamdata's figures, sales pipeline time fell by ten days while the stretch before it grew.
Research is the longest step. G2 says buyers have named research as the longest step in their journey for several years running, and for one in four the final decision is the hardest. The oldest widely quoted split of buying time comes from CEB, now part of Gartner, reported in 2018: buyers spent 27% of their time researching independently online, 18% researching offline and 17% meeting potential suppliers, which comes to about 5% per supplier when three are compared. That split predates the pandemic and AI search, so treat it as a ceiling on supplier time rather than a current figure.
Buyers want to do the research alone
Every recent survey points the same way: buyers prefer to get through most of the journey without a salesperson, and the preference is growing.
| Measure | Figure | Source and sample |
|---|---|---|
| Prefer a rep-free buying experience | 61% | Gartner, 632 buyers, August to September 2024 |
| Prefer a rep-free buying experience | 67% | Gartner, 646 buyers, August to September 2025 |
| Prefer a fully digital, self-service buying experience | 70% | Gartner, 2025, as stated in its May 2026 release |
| Prefer first sales contact at evaluation or decision | 62% (46% in 2024) | G2, 1,169 software buyers, April 2025 |
| Prefer first sales contact during research | 31% (43% in 2024) | G2, 2025 |
| Want to self-serve part or all of the journey | Virtually 100% | TrustRadius, over 2,000 technology buyers, 2022 |
| Buyer, not seller, initiated first contact | 79% | 6sense, 2025 |
Source: Gartner press releases of June 2025, March 2026 and May 2026, as republished; G2 Buyer Behavior Report 2025; TrustRadius 2022 B2B Buying Disconnect, via the TrustRadius vendor blog; 6sense 2025
G2 notes that its 2024 figures were re-weighted to remove a "discovery" stage, and that the shift towards later contact holds even if all discovery answers are counted as research. The TrustRadius figure is four years old and is included only because nothing newer measures the same question.
Rep-free does not mean rep-less. The same Gartner survey that found 67% preferring a rep-free experience also found 69% want to validate AI-generated insights with a salesperson, in a release of May 2026 covering 645 buyers. Buyers used an average of seven information sources in a recent purchase, and 45% used generative AI. Gartner also reported that 73% actively avoid suppliers who send irrelevant outreach. Read together, buyers want to arrive at the conversation with a shortlist already made, then use a rep to check it.
Which sources shape the software shortlist
G2's 2025 survey asked software buyers which external sources influenced their shortlist and which they trusted most at the decision. The vendor's own website sits in the top three for shortlisting, behind generative AI chatbots and software review sites. G2 runs a software review site, so its interest in this ranking should be stated.
| Source | Share of respondents |
|---|---|
| Generative AI chatbots | 17.1% |
| Software review sites | 15.1% |
| Vendor website | 12.8% |
| Market research firms | 10.6% |
| Independent content | 9.7% |
| Peers and colleagues | 8.9% |
| Vendor salesperson | 8.8% |
| Internal supplier portal | 8.1% |
| Independent forums | 7.6% |
Source: G2 Buyer Behavior Report 2025, survey of 1,169 B2B decision makers, April 2025
At the decision stage in North America, the order is similar: generative AI chatbots 17.2%, software review sites 13.4%, peers 13%, vendor sites 12.6% and market research firms 12.6%, with the vendor salesperson at 9.3%. G2 reports that 79% of buyers say AI search has changed how they research and 29% start research with AI search more often than Google.
Shortlists are small and fixed early. G2 found buyers prefer two or three vendors, and the share skipping a shortlist altogether has doubled in recent years. TrustRadius, now part of HG Insights, found 83% of 1,862 technology buyers shortlisted three or fewer products in its January 2026 survey, and 67% bought their first choice. 6sense found 94% ranked their shortlist before engaging any seller, and the first vendor contacted won roughly 80% of the time. Our B2B Website Statistics 2026 post covers the 6sense day-one shortlist figures in more detail.
Where the vendor website sits
The vendor website is one research source among about seven, and its weight rises with company size. Enterprise buyers use every source more, which reflects bigger buying groups and longer evaluations.
| Source | 1 to 250 staff | 250 to 1,000 | 1,000 to 5,000 | Over 5,000 |
|---|---|---|---|---|
| 47% | 48% | 51% | 53% | |
| Software review sites | 38% | 45% | 56% | 61% |
| AI search | 35% | 44% | 55% | 57% |
| Market research firms | 34% | 42% | 47% | 50% |
| Vendor sites | 33% | 45% | 48% | 50% |
| Peers and colleagues | 28% | 25% | 29% | 32% |
Source: G2 Buyer Behavior Report 2025
For an SMB-focused SaaS product, a third of buyers will use the website as a research source; for an enterprise product, half will. In both cases the review profile and the AI answer reach as many buyers or more, so the website has to agree with them.
Consistency is the problem buyers name. Gartner found 69% of B2B buyers report inconsistencies between the information on a vendor's website and what its sellers tell them. That figure matters more as the first sales call moves later: a buyer who has spent months on the website and then hears a different price, scope or integration story from the rep has reason to doubt both.
Buyers know what they want to find. The most recent TrustRadius survey we found that publishes the full list, from 2022, found buyers wanted vendors to provide pricing information (81%), product specifications (72%), free versions or trials (64%) and demos (44%). TrustRadius says transparent pricing has topped buyers' wish list for vendors every year since it began asking in 2023. Forrester's 2026 survey of nearly 18,000 business buyers adds that more than 60% bought some form of trial before committing, and just over a third planned to convert with the same provider.
The touchpoints are many, the conversations few. 6sense counted an average of 16 interactions between buyers and the vendor that won, one fewer than in 2024. Dreamdata, measuring every tracked touch, counted 88 across a journey, spread over four channels and ten stakeholders. Most of those 88 happen on owned and paid channels before anyone speaks to sales, which is why the website gets visited by people the CRM has never heard of.
What the data does not say
There is no recent, primary dataset on how many pages a B2B SaaS buyer views before requesting a demo, or what share of total research time is spent on vendor websites specifically. The CEB time split from 2018 is the closest, and it is old. Forrester's full reports, which hold more detail on digital interactions, are client-only; we have used only what Forrester publishes openly. Dreamdata's 2026 figures reached us through PPC Land's report of the release, not the report itself. The page-level numbers in our B2B SaaS Website Conversion Rate Statistics 2026 post are the best proxy for which pages carry the research stage.
What this means for a B2B SaaS site: a five-page test order
Put the numbers together and the brief is plain. Buyers spend most of a long journey researching alone, decide their shortlist of two or three before a rep calls, and check the website against review sites, AI answers and later the salesperson. The pages to fix first are the ones that answer research-stage questions without a call. This order follows the journey rather than traffic.
1. Pricing. Transparent pricing has been the top buyer request for four years. Publish a tier, a starting price or priced scenarios, and test the published version against "contact sales", measured on qualified pipeline rather than form fills.
2. Product and integration pages. Product specifications were the second most requested item. Each page should answer the questions an AI summary or a review site will be asked, in the same words, so the three sources agree.
3. Comparison and alternative pages. Shortlists of two or three are built from AI chatbots and review sites. A fair comparison page gives a buyer the same material on your site, and gives AI tools a source to cite.
4. Trial, tour or self-serve demo. With 70% preferring a fully digital process and more than 60% buying a trial first, a path to the product that does not require a meeting is the website's version of a first sales call.
5. Demo or contact page, last. It is the page buyers reach at the evaluation stage, by which point they have a ranked shortlist. Make it short and route it fast; the earlier pages decide whether anyone arrives.
We build marketing sites in Webflow on a four-week process, and this order is where we would start a B2B SaaS brief. Our marketing websites service page explains how we scope one.
