Lead scoring for SaaS: score stack, stage, and timing before the signal expires
Software buyers raise money, hire, and switch tools within months. Here is how to score them on fit and timing, with public data sources and a worked example.
Lead scoring for SaaS is the practice of ranking software and technology companies, and the people inside them, by how well they match the customers you win and how ready they are to buy. Software buyers raise money, hire, switch tools, and lose champions within months, so a model built for them weights tech stack, stage, and business model, and lets signals expire quickly.
This page is for vendors that sell dev tools, infrastructure, go-to-market tools, and finance and HR tools to startups and scale-ups. Scoring product usage in a trial or free plan is its own topic, covered in our PQL scoring guide; here it is one engagement input among several.
Why generic lead scoring fails when the buyer is a software company
A standard model scores up employee count, a senior title, and website activity. When the buyer is a software company, four things break it.
| What is different | What it does to a generic score | What to do instead |
|---|---|---|
| Signups arrive on personal email | The lead has no company, so fit is empty | Match the person to an account before scoring fit |
| Companies are young, and many shrink or close | Last year’s strong account may have pivoted or wound down | Refresh fit data every quarter |
| Total headcount hides the company’s shape | 120 engineers and 120 sellers score the same | Score headcount and hiring by function |
| The stack decides the deal | The account scores well and then cannot connect to your product | Make the stack the heaviest fit factor |
The second point shows up in public data. The Bureau of Labor Statistics reports that 29.1 percent of information-sector establishments born in March 2013 were still operating in March 2023, against 34.7 percent across the private sector. That sector is wider than software, but the direction holds.
Fit factors for software companies, and where the data comes from
CRM fit data is rarely complete. In Validity’s survey, 76% of CRM users said less than half their data is accurate and complete. Software companies are easier to research, because they publish pricing pages, documentation, and job postings.
| Fit factor | Why it matters | Where the data comes from |
|---|---|---|
| Tech stack (cloud provider, CRM, data warehouse, languages, the tools next to yours) | Decides whether you can integrate, and whether you replace or extend something | Job postings, public documentation, technographic providers, your own integration data |
| Stage and funding (bootstrapped, seed, Series A to C, growth, public) | Tells you the budget, the urgency, and whether a founder or a process approves | Funding databases, announcements, SEC Form D filings |
| Business model (product-led or sales-led, self-serve or contract) | Decides which problems they have and who buys | Their signup flow and pricing page |
| Headcount, growth rate, and team structure by function | Shows whether the role you sell to exists yet and is growing | Professional networks, careers pages, enrichment data |
| Geography and remote or office-based | Your coverage, data residency, and fit for IT and HR products | Headquarters and job posting locations |
Tech stack is the sharpest factor. List the systems you integrate with, the ones you replace, and the ones that rule you out, then score against that list. On sources, BuiltWith describes itself as detecting the technologies used on a website, such as analytics, frameworks, and hosting providers. HG Insights says its install data comes from content companies produce themselves, such as job postings, resumes, and the company’s website. A CRM or data warehouse is not visible on a public site, so check a provider’s coverage before you weight a factor on its data.
Funding has a public record in the United States. The SEC says a company selling securities under Regulation D must file a Form D notice within 15 days after the first sale, and that the Form D will be publicly available on EDGAR. The form asks for the date of first sale and the total amount sold. EDGAR’s full-text search covers filings since 2001 and can be limited to one form type. A filing confirms a round. A missing one tells you nothing.
Your ideal customer profile tells you whether stack, stage, or business model should weigh most. Our five-step framework for building an ICP scoring model covers the step from that profile to a set of weighted factors.
Lead scoring for startups
Revenue and total headcount say little about a 15-person company. Score funding recency, whether the role you sell to exists yet, and founder engagement. Consider a separate startup model with a lower fit bar and a cheaper play.
Timing and buying signals for software companies
Stack and stage decide which companies make the list. Timing decides which ones get worked this month.
| Signal | What it usually suggests | Where to find it |
|---|---|---|
| Funding round | New budget, a hiring plan, pressure to grow | Funding databases, announcements, Form D filings |
| First hire in your function (first RevOps, security, or finance hire) | The owner of your problem has just arrived | Careers page, professional networks |
| New executive (CRO, CTO, CFO, CISO) | The stack gets audited in the newcomer’s first quarter or two | Announcements, Form 8-K for public companies |
| Tool adopted or removed | A gap or a neighbor for your product has appeared | Technographic providers, job postings that name a tool |
| Product launch or pricing-page change (a new enterprise tier) | New workloads, or a shift in business model | Changelog, their pricing page |
| Layoffs | A spending freeze, often vendor consolidation | News, state WARN notice lists |
| Fiscal year end and renewal dates | Spend bunches at year end; a competitor’s renewal sets your window | Annual reports for public companies; discovery calls |
| Review-site research in your category | Someone is comparing products now | Your review-site vendor account, if you have that data |
Public companies. The SEC’s guide to Form 8-K says companies must make most 8-K disclosures within four business days of the triggering event. The items include officer departures and appointments (Item 5.02), completed acquisitions (Item 2.01), and exit or disposal costs (Item 2.05), where a restructuring appears.
Layoffs. Under the federal WARN Act, employers with 100 or more employees must give at least 60 calendar days of advance written notice of plant closings and mass layoffs. Some states publish the notices, such as California’s WARN reports. Score a layoff as a negative on timing, unless your product cuts cost or is the tool that stays in a consolidation.
Review sites. G2’s buyer intent data reference defines signal types that include Category, Profile, Pricing, and Compare. G2 describes a category page view as early-stage research and a comparison page view as direct product evaluation.
Set decay fast. Software buyers move quickly and lose interest as quickly. As a starting point, halve engagement points every 30 days, and treat a funding round or new executive as relevant for about two quarters. HubSpot’s scoring tool lets you set a percentage and a period for score decay. Funding rounds, first hires, and the other signals above each have a definition in our buying signals glossary.
Engagement: product usage, trials, and personal emails
For a software vendor, the strongest engagement is often inside the product. A trial started, an integration connected, and a teammate invited say more than an email open. Bring a few usage facts into the CRM as properties, such as active users at the account and integrations connected, and score them.
Many signups arrive on personal email addresses, and a rule that discards free-mail domains throws away real buyers. Handle them in three steps:
- Ask for the company at the point of value: a company domain or work email when they invite a teammate or connect an integration.
- Match the person to an account and associate the contact with the company record.
- Score fit only when matched. An unmatched signup has no fit data, and missing data contributes 0. Route it to self-serve until the match arrives.
Scoring the buying group
Software teams buy in groups too: Forrester reports that 95% of B2B purchases involve three or more people. In a software company the group is small, forms from the bottom as often as from the top, and changes with stage.
| Role | What their engagement tells you |
|---|---|
| User-champion (engineer, rep, analyst) | The product solves a real problem |
| Functional lead | There is an owner and a likely budget |
| Founder | The approver at a startup |
| Finance | A purchase is being sized |
| Security, at later stage | A formal review has started |
Coverage earns a factor to itself: count the functions, such as engineering, finance, and security, that have an engaged contact at the company. Six engineers is one function. Then weight by stage. At a seed company, the founder is most of the decision. At a 600-person company, a first contact from security or finance tells you more than another session from your champion.
Disqualifiers specific to software companies
- A stack you cannot work with: a CRM, cloud provider, or language you do not support.
- Too early. No team in your function and no plan to hire one. Rescore when the first hire appears.
- Students and job seekers, common in dev tool signups.
- Agencies and consultants evaluating for clients. Route them to a partner motion.
- Competitors. Keep a list of competitor domains and subtract on a match.
- Winding down or being acquired. The acquirer chooses the tools, so score the acquirer.
A personal email address alone does not disqualify a lead. Match it to a company first.
A worked example: scoring a demo software company
This demo account is a Series B software company, and the vendor scoring its fit sells a revenue operations tool. A recent layoff in the team that would buy is the only subtracting factor.
- Tech stack match (CRM and warehouse you support) 30% 8.7
- Stage and funding (Series B, raised 5 months ago) 25% 8.3
- Business model (sales-led, sells to mid-market) 20% 7.6
- Headcount and growth (40 sellers, 6 open sales roles) 20% 6.6
- Layoff in the buying team 5% -4.2
The math: 8.7 at 30% is 2.61, 8.3 at 25% is 2.075, 7.6 at 20% is 1.52, and 6.6 at 20% is 1.32, which adds to 7.525. The disqualifying factor subtracts 4.2 at 5%, or 0.21, for a fit score of 7.3.
The same account’s engagement over the last 60 days:
| Signal | Who or what | Counts toward engagement |
|---|---|---|
| First VP of Revenue Operations started six weeks ago | Functional lead, new executive | Strongly |
| Trial workspace opened, CRM integration connected | Product usage | Strongly |
| Three active trial users, one on a personal email matched to the account | User-champions | Moderately |
The account has a 7.3 fit, a new functional lead, and a connected trial. The layoff is a reason to ask about budget on the first call.
Plays by score band
Use these cutoffs to begin with, then move them as trial and deal data comes in.
| Engagement 6.0 and up | Engagement 3.0 to 5.9 | Engagement under 3.0 | |
|---|---|---|---|
| Fit 7.0 and up | Account owner acts within a day. Confirm the integration path and the budget owner | Reach out with a reason tied to an event: the round, the new leader, the open roles | Stay on the target list and look for a functional lead or founder to add |
| Fit 4.0 to 6.9 | Qualify stack, stage, and business model on a short call | Nurture by stage, and rescore after each quarterly data refresh | Rep time can wait |
| Fit under 4.0 | Check the data and the account match. If it holds, point them to self-serve | Self-serve and content only | Nothing to do |
Timing moves the plays:
- After a funding round or a new executive, work the account within 90 days.
- Late in the buyer’s fiscal year, unspent budget favors short deals.
- Before a competitor’s renewal, start early enough for an evaluation to finish.
- After a layoff, hold outbound for a quarter unless your product reduces cost.
How to measure and tune the model
The only test that counts is whether your highest-scored software accounts become pipeline and revenue at a higher rate than the others. Check conversion by band every month and retune every quarter.
| What you see | What it usually means | What to do |
|---|---|---|
| Bands convert at nearly the same rate | Headcount or funding total is carrying the model | Add stack and business model, then reweight |
| High-scoring accounts stall in technical review | Fit ignores integration | Raise the stack weight |
| Active trials score low | Signups are not matched to accounts | Fix the match step. Missing data adds nothing to a score |
If the bands still do not separate, our guide to where teams get ICP scoring wrong lists the common failures. Our account scoring worksheet is a blank sheet for laying out stack, stage, and model factors.
Where BoostIdeal fits
BoostIdeal is a scoring engine for teams that run their software-company pipeline in HubSpot.
- You decide the logic. A person builds the model and sets what stack, stage, and business model are worth. AI helps analyze your trial and deal data and draft a first pass.
- Any record can be scored: accounts, contacts, leads, and custom objects. One record can be scored by another, such as a contact by its account’s fit.
- Research comes in through HubSpot or CSV upload, so a funding date, a stack field, or an active-user count becomes a property you score on.
- Each score lists what drove it. Scores are out of 10, and the driving factors write back to HubSpot properties, ready for your workflows, lists, and agents.
- Startups and later-stage companies can be scored by separate models running side by side.
- Fast-moving accounts need fresh scores: they recalculate hourly by default, and in near real time with webhook sync.
When it is not the right fit. Skip it if your trial signups and accounts sit outside HubSpot and exporting them to CSV for upload is not something you want to do, or if you would rather a vendor hand over a number you never question.
Frequently asked questions
What is lead scoring for SaaS?
It is how teams that sell to software companies rank those companies and their contacts by fit and buying readiness. Fit comes from tech stack, stage and funding, business model, and headcount by function. Readiness comes from events such as funding and first hires, product usage, and buying group coverage.
What should a B2B SaaS lead scoring model include?
Two scores kept apart: fit and engagement. Fit covers company attributes, with tech stack usually the heaviest. Engagement covers what people at the account do, including trial and product usage, with fast decay. Add subtracting factors for students, agencies, and competitors.
How do you score leads at startups?
Score funding recency, the first hire in the function you sell to, and founder engagement. Revenue and total headcount say little at this size.
Should I use lead scoring or account scoring for SaaS?
Use both, with the account first. Fit belongs to the company: its stack, stage, and business model. Contact scores tell you who to reach inside a strong account.
How should I score signups that use a personal email address?
Match them to a company before scoring fit. Ask for a company domain or work email when the user invites a teammate or connects an integration. Until then, missing fit data contributes 0, so route the lead to self-serve.
Is this the same as PQL scoring?
No. PQL scoring ranks users and accounts by what they do inside your product during a trial or free plan. The model here ranks software companies by fit and buying signals, with product usage as one input.
The takeaway
Software companies change faster than your CRM does. Put the stack at the center of fit, add stage, business model, and function-level headcount, and let funding, first hires, tool changes, and product usage set the timing. Match personal-email signups to accounts, set decay fast, and never hide a factor or a weight, so a rep can see why an account ranks where it does.