Tech-Stack and Traffic Fit
Outcome: stack and traffic columns on qualified accounts, used as positive qualifiers and score inputs — never as exclusions on a negative.
- Surface
- App and MCP server
- Level
- Intermediate
- Uses
- find_company_techstack · find_company_website_traffic
- Credits
- ~1 per company across both
- Prerequisite
- Lesson 04's qualified accounts
Run on survivors only
Both columns in this lesson cost more than the firmographic bundle, so they run after lesson 04’s filter, on qualified rows only. On a universe that halved at the firmographic stage, that is half the cost for the same result.
What stack detection sees
Detection works from what a site exposes publicly — scripts, DNS records, response headers, embedded widgets.
Strong coverage — the front end
Analytics, tag managers, chat widgets, marketing automation, CDNs, payment providers, CMS, ad pixels. These load in a browser, so they are reliably visible.
Weak coverage — the back end
CRM, data warehouse, internal tooling, databases, HR systems. A company running Salesforce internally may expose nothing about it publicly.
No coverage — anything private
Custom internal systems, anything behind auth.
The asymmetry produces the single most important rule in this lesson.
A positive detection is strong evidence. A negative is not evidence at all. Filtering out companies where a tool was not detected removes accounts on the basis of a public-website limitation — and it disproportionately removes larger, more security-conscious companies, which are often the ones you most want. Use detection to promote, never to exclude.
Three ways to use the stack
| Use | How | Strength |
|---|---|---|
| Competitor displacement | They run a competing tool → high priority | Strongest single qualifier most teams have |
| Complement | They run a tool yours works alongside → good fit | Strong |
| Maturity proxy | Number and sophistication of detected tools → they invest in tooling | Moderate |
The first is worth building the workflow around if it applies to you. A confirmed competitor account with a timing signal is the highest-value row in most B2B pipelines.
Traffic as a size proxy
find_company_website_traffic gives monthly visits, trend and channel mix. Two honest uses:
- Size, where headcount is unreliable. E-commerce, media, consumer and marketplace businesses are often small in headcount and large in reach. Traffic is the better size signal for them.
- Trajectory. The trend matters more than the number. Traffic up 40% over six months is a signal worth acting on; 200k monthly visits on its own is just a fact.
Deeper keyword and channel analysis lives in SEMrush and Similarweb if those connections are available.
Combining into a fit view
At this point every qualified account has firmographics, stack and traffic. The combination is what separates a list from a market map:
| Pattern | Reading |
|---|---|
| Right size + competitor detected | Displacement target — highest priority |
| Right size + complement detected | Natural fit |
| Right size + high tool count | Invests in tooling; likely to evaluate |
| Right size + traffic growing fast | Scaling — pain is arriving |
| Right size + nothing detected | Unknown, not disqualified |
That last row is the point. Roughly a fifth of accounts will show nothing, and they belong in the pipeline with a neutral score rather than in the bin.
Do this now
Filter to qualified accounts
From lesson 04. Confirm the count before running.
Run stack detection on five rows
Read the detected tools. Do they match what you would expect for companies you know?
Widen to the qualified set
Build the three usage columns
competitor_detected, complement_detected, tool_count.
Add traffic where it is meaningful
Only for segments where reach is a better size proxy than headcount.
Build the fit view
Combining firmographics, stack and traffic patterns.
Confirm no negative exclusions
Check that no filter drops rows for a non-detection.
Check your work
- Both columns ran only on qualified accounts
- Positive detections are used to promote, not negatives to exclude
- Accounts with nothing detected carry a neutral score, not a zero
- Traffic is used only where it is a better size proxy than headcount
- You spot-checked detections against companies you know
Where this breaks
Treating “no competitor detected” as “not a displacement target” systematically removes your best accounts. Back-office and internal tools are the least visible category in public-site detection, and they are exactly the tools most B2B products displace. If your product replaces something that does not render in a browser, stack detection cannot qualify your market — use job posts instead, which name internal tools explicitly. See Hiring Signals.
Further automation
Stack changes are signals. A company that adds a competing tool has just entered a buying cycle; one that removes it has just left a vendor. Re-running detection quarterly and diffing gives you both, and Techstack change is built for exactly that.
Next lesson
06 — Find the buying committee, moving from accounts to people.
Reference for this lesson: find_company_techstack, find_company_website_traffic, Company techstack, Techstack change.