Source: Growth Unhinged, one of the top GTM newsletters for B2B SaaS founders and revenue leaders. Everything below is current as of Q1 2026.
1. AI for GTM: Plays That Are Working
Content and Competitive Intelligence Plays
AI Content Assistant
Set up a Claude or ChatGPT Project. Upload your brand guidelines, product descriptions, and 5 to 10 examples of your best-performing content. Write detailed project instructions covering your voice, and your dos and don'ts. Use it to ideate, draft, and edit LinkedIn posts across the team — anyone on the GTM team can produce on-brand content without writing a full brief each time.
Competitor Comparison Pages at Scale
Stack: Clay + Claude + Webflow + Make.
- Build lists of competitors across market categories in Clay.
- Enrich via Clay: scrape website, features, reviews, and pricing data.
- Run Claude columns to generate comparison copy, strengths and weaknesses, and FAQs.
- Format the copy to HTML and push it to Webflow via Make.
One team shipped 1,600+ pages in 2 months this way, with triple-digit growth in organic traffic and a #1 ranking for competitor pricing queries in AI Overviews.
AI Digital Twin for Customer Research
Feed Claude real customer data: call transcripts, G2 reviews, and CRM objection notes. Ask it to surface patterns in motivations, objections, and buying triggers. Create 5 anchor statements ranging from "too complex, not worth switching" to "exactly what we need," and load them into a Claude Project as a persistent digital twin. Test any new campaign or message against it before launch.
This only works with real customer data — synthetic data in, synthetic insights out.
AI Competitive Intelligence Copilot
Run daily automated research pipelines on competitors: news, call data, open opportunities. AI does a first pass of analysis, a human approves it, and the intelligence is surfaced to sellers via a chat interface. Sellers get real-time competitive answers tailored to specific deals instead of digging through static battle cards.
Prospecting and Outbound Plays
AI SAM Scoring Algorithm
- Create a true/false output field in your CRM, plus a reasoning field.
- Define non-negotiables (firmographics: revenue, industry, size) that any account must pass first.
- Define nice-to-haves (job titles, tech stack, recent signals) for the AI scoring layer.
- Build a Claygent column with a research prompt that outputs true or false.
- Test on 20 to 60 known accounts before running at scale.
- Backfill existing CRM accounts and auto-score new ones on entry.
Only run the AI scoring step on accounts that already pass the non-negotiables — it keeps costs down.
AI Personalized ABM at Scale
Use AI to scan prospect websites: pull their logo, brand colors, and website messaging, then generate a sample of your product's output customized to their company. Automate the whole thing via n8n + OpenAI + Apollo.
Cost: under $10 in AI credits for 1,000+ personalized emails. One team booked more meetings per month with this system than they did with two full-time BDRs.
AI Meeting Prep
Build a custom GPT or Claude Project. Before every call, paste in: calendar event details, the email thread, prior meeting transcripts, the LinkedIn conversation, company details, and the other party's LinkedIn profile.
Output: an attendees overview, meeting purpose, summary of past meetings, and prospect context. Time to set up: 1 hour. Time saved per call: 20+ minutes.
Sales and Lifecycle Plays
AI Inbound BDR
Deploy an AI BDR to handle personalized outreach sequences for new signups, knowledge-base-powered responses to inbound questions, and meeting booking synced with AE calendars. Pair it with AI-powered waterfall enrichment running across multiple data providers in parallel for better coverage.
SafetyCulture's result running this play: 3x meeting bookings, 2x opportunities created.
AI Re-engagement for Closed-Lost Deals
Trigger: a deal moves to closed lost in the CRM. Wait 90 days, then run an AI research agent to find new marketing leadership, recent campaigns, major announcements, and GTM motion changes. Generate personalized content per contact: 4 marketing emails, 4 sales emails, 2 LinkedIn DMs, and 1 custom landing page.
Closed-lost deals are easier and cheaper to re-engage than cold outreach to new logos.
2. Free-to-Paid Conversion
2026 Conversion Benchmarks
Data from 200 B2B software products, January 2026.
What good looks like
| Model | Good | Great |
|---|---|---|
| Freemium | 3–5% | 8–12% |
| Free trial (no credit card) | 4–6% | 10–15% |
| Free trial (credit card required) | 25–35% | 50–60% |
| AI-native products | 6–8% | 15–20% |
Per 1,000 website visitors
| Model | Signups | Paying customers |
|---|---|---|
| Freemium | 90 | 5 |
| Free trial (no CC) | 45 | 3.6 |
| Ungated experience | 70 | 5.6 |
| CC-required trial | 35 | 10.5 |
Credit-card-required trials produce the most paying customers per 1,000 visitors, even with lower raw signup numbers.
Freemium vs. free trial is the wrong question
If your goal is adoption: open up the product before requiring an account (an ungated experience). If your goal is conversion: require a credit card to start a trial. The worst outcome is a timid free experience that neither drives adoption nor creates urgency to buy.
7 Tactics That Improve Conversion
- Require a credit card. The single highest-impact conversion lever in the dataset. Fyxer tested this and conversion jumped from 5% to 35% — paying customers doubled even though signups dipped. Make it feel safe: show a timeline of what happens today, at day 5, and at day 7, and send a reminder email before the card is charged.
- Default to annual pricing. Set annual as the default plan, offer a 25% discount, and show the effective monthly price. One team saw 2.3x more new trials choose annual, and 50% of their paying customers are now on annual plans.
- Add a dual CTA. Instead of one "get started for free" CTA, add a second for a premium trial. One team saw a 26% improvement in premium trial signups from this change alone.
- Segment trial lengths by user type. Work email signups get a 7-day trial; personal email signups get a 14-day trial, since they take longer to reach the aha moment. Segmenting increased personal-user trial start rate from 13.4% to 22.1%.
- Invest in LLM optimization (AEO). ChatGPT traffic converts at 24% at Webflow — 6x higher than Google. LLM-sourced signups are the highest-converting acquisition channel right now. What's working: free template libraries (one team reported 40x SEO impact), G2 review investment (LLMs frequently cite G2), technical SEO experiments to improve how pages are parsed by AI search, and localized landing pages.
- Fix the activation ownership gap. Product owns activation 49% of the time, but conversion accountability falls to sales or growth. That disconnect means activation never gets the attention it needs — assign one person accountable for both activation and conversion.
- Move team invites post-paywall. Move the team invite prompt to after the paywall, pre-populated with suggested team members. One team sees 1 in 3 invites accepted. Limit referral rewards to work email signups to reduce program abuse.
3. Pipeline From Personal Emails
The Opportunity and the 4-Step Play
The problem
AI-native products see 75 to 90% of signups from personal emails — bolt.new sees 98% personal email signups. Most GTM teams either block personal emails or ignore them, and both approaches leave pipeline on the table. Why it's happening: users trial AI tools personally to avoid procurement, then bring them to work. The line between personal and professional AI use is blurring fast.
The 4-step play
- Don't block personal email signups — keep the funnel open.
- Run personal email signups through an enrichment tool (e.g. Freckle) to identify which users work at companies in your ICP.
- Route identified users into targeted GTM plays: personalized outbound sequences, ABM campaigns, or LinkedIn ads.
- Add a product path prompting personal users to also connect their work email.
Why it works now
Enrichment accuracy on personal emails has had a step-function improvement in the past year, and the economics now make sense at scale. Work email signups have 10x higher LTV than personal email signups, so even a low enrichment match rate produces high-value pipeline.
bolt.new's result: $1.7M in B2B pipeline in the first 4 weeks, with 23% of their B2B pipeline now coming from personal email users.
4. Fyxer: 514 Experiments in a Year
Experiments Worth Stealing
Fyxer ran 514 experiments in a year. These are the ones worth stealing.
- Credit card required trial. Conversion jumped from 5% to 35%. Total paying customers doubled even though raw signups dipped — Fyxer called it a winner after 8 days of testing. Show a clear timeline of what happens today, at day 5, and at day 7, and send an email reminder before the card is charged.
- Annual default with a 25% discount. Set annual as the default plan, offer 25% off, and show the effective monthly price. This 2.3x'd the share of new trials choosing annual — 50% of paying customers are now on annual plans.
- Price increase via a new tier. Introduced a Pro tier at $50/user/month, up from $30. Month-0 revenue per trial increased by 67%, while checkout rate dropped by only 6%. The key: introduce a new feature tier at the higher price point rather than raising the price on the existing plan.
- Trial length segmentation. 7-day trial for work email signups, 14-day trial for personal email signups. Personal user trial start rate went from 13.4% in the control to 22.1% in the treatment — a 65% increase.
- Team invites post-paywall. Moved team invite prompts to after the paywall, pre-populated with the user's closest team members. 1 in 3 invites are accepted. The referral program is now restricted to work email signups to reduce abuse.
The Targeting Insight
Work email signups have 10x higher LTV
Fyxer enriched all signup emails via Apollo and analyzed expected LTV by signup type. Work email signups had 10x higher LTV than personal email signups. They didn't shut off personal signups — instead they revised all marketing to attract and optimize for work email signups, and added a product path for personal users to also connect their work email.
What this means for GTM teams
Most free funnels treat all signups as equal. They aren't. Enriching your signup list and segmenting by ICP fit before running any growth experiment means your conversion improvements compound on a better base. Get the data right before you start experimenting — the targeting insight comes first.
5. AI Pricing Models Worth Studying
The Platform Plus Tokens Model
The problem with pure credit-based pricing
Credit-based pricing conflates two things: the value you deliver (platform) and the cost of AI infrastructure (tokens). Buyers can't optimize their spend because they don't control what drives cost, and vendors have no margin floor if power users consume heavily.
The platform + tokens split
Separate pricing into two tracks:
- Platform: a high-margin subscription for the unique value you deliver. Customers pay this regardless of AI usage.
- Tokens: a cost pass-through with roughly 20% markup. Customers benefit directly when LLM costs drop.
Think of it as paying for a car lease (platform) and then paying for fuel as you drive (tokens).
Why it works
It guarantees a margin floor — no more unprofitable power users. Customers are incentivized to optimize token spend, which frees up budget for the higher-margin platform. It separates your unique value from commodity AI infrastructure, and it opens the door to bring-your-own-key and AI marketplace distribution models.
The diagnostic question
If your AI infrastructure costs went to zero tomorrow, what would be the fairest way to charge for your product? That answer is your platform pricing. Everything else is tokens.
Examples in the Wild
PostHog
AI features as a straight pass-through: actual AI token costs plus a 20% markup, with every customer getting $20 of free usage to start. PostHog can do this because their value lives in 10+ other products — AI features make those products stickier, not the monetizable product itself.
Clay (March 2026)
Formally split pricing into two axes: data credits (tokens) and platform features (actions). Data credit costs dropped 50 to 90%, with zero markup on AI model token costs. Clay accepted a 10% short-term revenue hit to open up platform adoption at scale — the bet is that cheaper data drives more complex platform usage, which carries higher margin.
Figma (March 2026)
Introduced an AI credit model in December 2025 and enforced it in March 2026, after 3 months of free usage used to collect consumption data and identify power users. 75% of customers with $10,000+ ARR are using AI credits weekly. Credits are allocated per user, with pooled top-up subscriptions available for teams that exceed limits.
