# SaaS marketing trends > Nine shifts with evidence, including AI answer engines as a referral source, rising CAC, the retreat of the MQL and the move from seats to usage pricing. Source: https://saas-marketing.net/guides/saas-marketing-trends/ Topic: SaaS Marketing Type: listicle Published: 2026-09-11 Last updated: 2026-09-11 Publisher: SaaS Marketing (saas-marketing.net) License: CC BY 4.0. Quote or republish with attribution and a link to https://saas-marketing.net/guides/saas-marketing-trends/ ## Short answer Nine shifts have genuinely changed B2B SaaS marketing practice: AI answer engines becoming a referral source, AI Overviews cutting organic click through rate, acquisition costs rising sharply since 2023, buying groups self educating before contact, the MQL losing its routing role, seat pricing giving way to usage and outcome pricing, efficiency metrics replacing growth targets, dark social becoming decisive but unmeasurable, and content being judged on citation rather than sessions. Only two require a budget change. ## Key takeaways - AI referred sessions grew sharply through 2025, but they arrive in tiny volumes with unusually high intent and convert well. - AI Overviews appear on roughly 48% of queries and cut organic click through rate by around 61% where they show. - B2B SaaS customer acquisition cost has risen materially since 2023, partly through real competition and partly through attribution loss. - The MQL is losing its routing role to buying group signals and product usage, not because it was wrong but because it was singular. - Seven of these nine shifts are measurement changes, not budget changes, which is why teams keep deferring them. --- Most trend posts are a list of things that might happen written by someone with no exposure to whether they do. This one only includes shifts with evidence behind them, each marked with a confidence level, and it says which ones need money versus which ones only need you to change what you count. Seven of the nine are measurement changes. That is the actual story. ## 1. AI answer engines became a real referral source (high confidence) Small volume, unusual quality. Similarweb reported AI referred sessions growing 527% year on year through mid 2025, and Google Search Console now reports generative surfaces separately, so this stopped being a thing you infer from weird direct traffic. The volume is still tiny for most B2B SaaS sites, often under 3% of sessions. The conversion behaviour is not. Someone arriving from a ChatGPT answer that recommended three tools has already done comparison work, so they land deeper in the funnel and convert at rates that look like branded search rather than cold organic. **What to change this quarter:** segment AI referrers in analytics, and stop letting them sit in direct. **Budget change:** none. ## 2. AI Overviews broke the traffic equals success assumption (high confidence) AI Overviews now appear on roughly 48% of queries, and around 82% of B2B technology queries, with organic click through rate falling about 61% where they show. Rankings held, traffic fell, and a lot of good content programs got declared failures in Q4 board reviews on that basis. The job of a blog post changed. It now supports branded search, feeds sales conversations, earns citations in answers, and passes internal link equity to pages that convert. Sessions are a weak proxy for all four. If your content dashboard's headline metric is organic sessions, you will make the wrong call in the next budget cycle. Swap the headline for branded search volume, citation appearances and content assisted pipeline before somebody uses a traffic chart to cut the budget that produces them. **What to change:** rewrite the content report. **Budget change:** none. ## 3. Acquisition costs rose 40 to 60% since 2023 (medium to high confidence) Public and private benchmark sets broadly agree on direction and roughly on magnitude, though methodologies differ enough that the precise figure should be treated as a range rather than a fact. Part of this is real: more funded competitors bidding on the same intent keywords, larger buying committees, longer cycles. Part is measurement. Attribution loss after browser and platform privacy changes moved credit away from organic and dark channels toward the last click that could still be tracked, which is usually paid. So measured CAC rose faster than actual CAC. Both halves matter and they need different responses. Rebuild your budget assumptions on the new numbers rather than last year's, which is the entire point of the [SaaS marketing budget calculator](/calculators/marketing-budget/) and the [SaaS marketing budget template](/templates/saas-marketing-budget/). **Budget change:** yes. This is one of the two. ## 4. Buying groups build shortlists before contacting anyone (high confidence) The consistent finding across Gartner and 6sense buyer research over several years is that B2B software buyers complete most of their evaluation independently, and often arrive with a shortlist already formed. By the time a form gets filled, the decision is frequently between two vendors rather than ten. The operational consequence is that your comparison pages, your G2 profile, your pricing transparency and your documentation are doing sales work in rooms you are not in. A product that hides pricing gets cut from shortlists silently, and nobody tells you. **What to change:** build the three pages that serve non champions. **Budget change:** minor, mostly writing time. ## 5. The MQL lost its role as the routing trigger (medium confidence) Not dead, demoted. A single person filling a form is still worth knowing about, but it stopped being sufficient for routing when six to ten people evaluate the purchase and the one who fills the form is often the most junior. Teams are shifting to account level or buying group qualification, blending fit, multi contact engagement and product usage. The organisations doing this well are not the ones with the best model; they are the ones who got sales to agree on a definition and then actually validated it against closed won data. **What to change:** add account level engagement to routing. **Budget change:** none, though it costs ops hours. ## 6. Seat pricing is under pressure where AI does the work (high confidence) The logic is simple and hard to argue with. If your product replaces work a person did, the customer succeeding means they need fewer seats, so your revenue falls as your value rises. Support and SDR tooling feel this first. Seat pricing is not dying generally. Figma, Slack and Notion charge per seat because more people genuinely does mean more value, and nothing about AI changes that. What is changing is that AI native products are landing on hybrid commit plus overage, in the shape Snowflake and Twilio established years ago, rather than pure per user. **82%** Share of B2B technology queries that trigger an AI Overview, the highest of any commercial category **What to change:** check whether your value metric still tracks customer value. **Budget change:** no, but a pricing project is real work. ## 7. Efficiency metrics replaced growth targets in board decks (high confidence) CAC payback, net revenue retention and burn multiple now sit above growth rate in most board packs, which is a durable change rather than a downturn artefact. It has been true since 2023 and has not reverted. For marketing this means the defensible budget request has an efficiency argument attached. "We will produce 40% more MQLs" lands badly now. "This moves CAC payback from 21 to 17 months, here is the arithmetic" lands. Build the request in those terms, which is how the [SaaS marketing plan template](/templates/saas-marketing-plan/) frames it. ## 8. Dark social became decisive and remains unmeasurable (medium confidence) Private Slack and Discord communities, podcasts, LinkedIn feeds, peer conversations. None of it passes a referrer, all of it influences shortlists. The practical response is self reported attribution: a free text or dropdown "how did you hear about us" field on the demo form. It is imprecise, biased toward recent memory, and still more informative than a platform dashboard claiming Reddit sent you eleven sessions when your sales calls keep mentioning a Reddit thread. [SaaS marketing on Reddit](/guides/saas-marketing-reddit/) covers the tactical side of the channel that most often shows this gap. Report platform attributed pipeline and self reported source in the same table every month. The gap between them is the most useful thing on the page, and it prevents a board from cutting a channel because a tracking system cannot see it. ## 9. Content is judged on citation and internal link support, not sessions (medium confidence) This follows from shifts one, two and four, and it is the one most teams have not operationalised. A glossary definition that ranks nowhere but gets quoted by three models, and passes authority to a comparison page that converts at 7%, is doing more work than a 3,000 word guide with 900 monthly sessions and a 0.4% conversion rate. Measuring it is awkward. Manual prompt testing across ChatGPT, Perplexity and Gemini on your top twenty buying queries, once a quarter, is crude but workable and takes about three hours. ## What this means for next quarter Two of the nine need budget: rising acquisition costs, and pricing work if your value metric has drifted. The other seven are measurement and process changes that cost ops hours and political capital rather than money, which is precisely why they keep getting deferred to a quarter that never comes. **The nine, as a work list** Pick two. Nine simultaneous changes is how a quarter ends with nine half finished projects. For the underlying structure these shifts act on, [the SaaS marketing funnel](/guides/saas-marketing-funnel/) and [how to scale SaaS marketing](/guides/scale-saas-marketing/) are the right next reads, and if your ICP definition predates 2024 it is worth redoing with the [ICP template for SaaS](/templates/ideal-customer-profile/) before anything else. Wider context sits in the [SaaS marketing](/saas-marketing/) hub and in [marketing software as a service](/guides/marketing-software-as-a-service/). ## Frequently asked questions ### Does content marketing still work for SaaS in 2026 with AI Overviews? Yes, but the job changed. AI Overviews appear on roughly 48% of queries and around 82% of B2B technology queries, and click through rate falls sharply where they appear. Content now earns citations, supports branded search, and feeds sales conversations rather than delivering raw sessions. Teams still measuring blog success by traffic alone will conclude content stopped working when it simply stopped being measurable that way. ### How do you get a SaaS brand cited by ChatGPT and Perplexity? Publish definitional and comparative content with clean structure: a direct answer in the first two sentences under each heading, tables where comparison happens, named sources for every number, and consistent entity naming across the site. Third party mentions on sites the models already trust matter as much as your own pages. Vague thought leadership essays are almost never cited. ### Why has SaaS customer acquisition cost risen so much? Three causes compound. Paid auction prices rose as more funded companies bid on the same intent terms. Attribution loss after privacy changes means some acquisition that organic earned now gets credited to paid, inflating measured CAC. And buying committees grew, so each deal requires more touches across more people before anyone talks to sales. ### Is the MQL dead? It is demoted, not dead. A single contact hand raise is still a useful signal, it just stopped being sufficient for routing when six to ten people evaluate the purchase. Teams are moving to account or buying group level qualification, with product usage signals carrying more weight than form fills. The MQL survives as one input rather than the trigger. ### Should SaaS companies move from seat pricing to usage pricing? Only where usage is the value metric. Seat pricing is fine for collaboration products where more people genuinely means more value, and Figma and Slack are not in trouble. The pressure is on products where AI does work a person used to do, since seat counts then fall as the product succeeds. Most such companies land on hybrid commit plus overage rather than pure usage. ### What is dark social and how do you measure it? Dark social is influence that leaves no referrer: private Slack communities, podcast listens, LinkedIn feed exposure, word of mouth between peers. You cannot track it, so you measure it by asking. A self reported attribution field on the demo form consistently surfaces channels platform data shows as near zero, and the gap between the two is the thing worth reporting.