Signal-Based Selling: Replacing the MQL With Buying Signals
Why the MQL stopped working
The marketing qualified lead was built for a world where a whitepaper download meant interest. Hit a points threshold, become an MQL, get handed to sales. In practice, many sales teams learned to ignore the queue: students downloading ebooks, competitors researching, and real buyers who never filled in a form at all because they did their research in AI tools, peer communities and review sites.
Signal-based selling flips the question. Instead of "has this person engaged with our content enough?", it asks "is there evidence that this account has a problem we solve, right now?" Timing becomes the main input, and volume becomes secondary.
The three families of signals
First-party signals
Things you observe directly, and the most reliable ones you have:
- Visits to high-intent pages: pricing, integrations, security, comparison pages.
- Product usage in a free trial or freemium plan: invites sent, key feature activated, usage limit hit.
- Repeat visits from several people at the same company within a short window.
- Replies, meeting bookings and support tickets that mention buying topics.
Second-party signals
Data shared by a partner: review site activity on your category page, event attendance from a co-hosted webinar, marketplace listing views.
Third-party signals
Public or purchased data about change at the account:
- Hiring: a company posting several roles that use your kind of tool.
- Leadership change: a new head of marketing or sales in their first months, often reviewing the stack.
- Funding or expansion: new budget, new offices, new markets.
- Technology change: adopting or dropping a complementary or competing product.
- Topic intent: a surge in research on topics related to your category across publisher networks.
Turning signals into a score
A single signal is noise. Signals that stack on the same account inside a short window are a pattern. A practical model has three parts:
- Fit: does the account match your ideal customer profile (industry, size, region, tech stack)? This is a gate, not a bonus. A perfect signal from a poor-fit account is still a poor lead.
- Signal weight: each signal type gets a weight based on how often it preceded a closed deal historically. A pricing page visit by three people usually outweighs one ebook download.
- Decay: signals lose value over time. A common approach is exponential decay, where a signal's weight halves every set number of days. Last week's pricing visit matters; last quarter's barely does.
Expressed simply: account_score = fit_gate × Σ (weight_i × 0.5^(age_days_i / half_life)). Start with weights set by sales judgement, then replace them with weights learned from your closed-won and closed-lost history once you have enough data.
Routing: what happens when the score fires
A signal is only useful if someone acts on it quickly. Define plays for each pattern:
- Hot inbound (pricing visits plus a known contact): assign to a rep within the hour with context attached.
- Warm account, unknown visitors: identify likely buyers at the account and start a short, relevant sequence that references the problem, not the website visit.
- Trigger event (new leader, new funding): send a tailored note about what teams typically prioritise after that event.
- Product-qualified (trial user hits a limit): offer help or an upgrade conversation from someone who understands their usage.
Never open with "I saw you were on our pricing page." It feels like surveillance. Use the signal to choose who and when, and let the message be about their situation.
Where AI fits
AI has made signal-based selling practical for small teams. Language models can read job posts, press releases and company pages and classify them ("this posting suggests they are building an outbound team"), something that used to need a research analyst. They can summarise the full signal history of an account into three lines a rep reads before a call, and draft a first message grounded in the actual trigger. Machine learning models trained on your own deal history can then tune weights and half-lives far better than a spreadsheet.
The rep still decides. AI ranks and explains; people choose the angle and build the relationship.
Getting started without a big budget
- Instrument your high-intent pages and send those events into your CRM against the account.
- Pick three signals you can already capture and agree weights with sales.
- Build one view: accounts ranked by score this week, with the reason shown.
- Review results monthly: which signals preceded real pipeline? Adjust.
A team at Adasify Studios could run this from a CRM, a web analytics feed and a shared dashboard before buying any intent data at all.
Key takeaways
- Signal-based selling prioritises timing and evidence of need over content engagement points.
- Combine first-, second- and third-party signals, and gate everything on ideal customer fit.
- Weight signals by history and let them decay over time.
- Route each signal pattern to a defined play, and never mention the tracking in outreach.
- AI classifies, summarises and ranks; people sell.
