Competitive Intelligence Playbook: A 5-Step Weekly Framework for B2B SaaS Teams

Most competitive intelligence programs fail for the same reason most workout routines fail: they're too ambitious on Day 1 and too easy to skip by Week 3.

The founder swears they'll check competitor pricing every Monday. By Wednesday they've checked it zero times. The VP of Product wants a monthly landscape deep-dive that everyone on Slack swears they'll read. Nobody reads it. The PMM builds a Notion database with 47 competitor fields across 12 tabs. It is updated twice — both times by the PMM, both times the week the board asked for a competitive update.

This isn't a motivation problem. It's a framework problem. The CI programs that stick aren't the most comprehensive. They're the most repeatable.

Here's the playbook. Five steps. Ninety minutes a week. No dedicated analyst required. It catches roughly 80% of the competitive signals that actually change how you should think or act. The remaining 20% — the deep strategic analysis, the full win/loss synthesis — you do quarterly. Weekly is for signal detection. Quarterly is for strategy.

Step 1 — The 10-Minute Pricing Scan (Monday, 8:00 AM)

Pricing changes are the single highest-signal competitive event in B2B SaaS. When a competitor changes pricing, they're telling you something about their strategy: their unit economics, their target segment, their desperation level. A price increase usually means they're confident about retention. A price cut usually means they're not.

Your job every Monday: open five competitor pricing pages. Look for three things and three things only:

• Numbers that changed. Any tier price, seat minimum, or usage limit that moved since last week. Don't analyze — just flag it.
• Plans that appeared or disappeared. A new "Startup" plan means they're moving downmarket. A vanished "Free" tier means cash is tight. A new "Enterprise" tier means they landed a big customer and are building the plan around them.
• Feature gates that shifted. Did SSO move from Enterprise to Pro? Did API access get rate-limited? Feature gates are the quiet pricing lever — they change what you get for the price without changing the price itself.

Ten minutes. Five competitors. Three questions each. Write down anything that changed. Don't write a narrative. Just note the delta. That's the whole step. For teams tracking more competitors or dealing with JavaScript-heavy pricing pages, here's how to automate the pricing scan entirely.

Step 2 — The 15-Minute Hiring Pulse Check (Monday, 8:10 AM)

Hiring is the most underrated competitive intelligence signal in SaaS. It's also the most predictive — job postings appear 3 to 12 months before the resulting product changes hit the market. By the time a competitor launches a feature, the hiring signal that preceded it is stone cold. You want the hiring signal, not the launch announcement. We've written a full guide on reading hiring signals, but here's the weekly checklist:

• New departments. Is a competitor suddenly hiring ML engineers when they've never had an ML role before? That's not a side project. That's a strategy change.
• Headcount velocity. Is one department hiring faster than last month? Sales hiring means they're scaling revenue. Engineering hiring means they're building. If both are accelerating, they just raised money or hit a growth inflection.
• Geography shifts. A competitor opening their first EMEA sales role is a geographic expansion signal. Five EMEA roles in one week is a geographic expansion commitment.

Check job boards — Greenhouse, Lever, Ashby, LinkedIn — for your competitor set. You don't need to read every job description. Scan for department counts and new role titles. Fifteen minutes. Flag anomalies. Move on.

Step 3 — The 20-Minute Review Sweep (Monday, 8:25 AM)

Customer reviews are a real-time sentiment stream. But reading 200 reviews a week is a recipe for quitting your CI program by February. You need a triage system.

Open G2 and Capterra for each competitor. Sort by most recent. Read the 5 most recent reviews — not the featured ones, the most recent ones. The featured reviews are curated. The recent ones aren't. Look for:

• Clustered complaints. Three reviews in two weeks mentioning the same bug or missing feature? That's not noise. That's a product weakness your sales team should know about.
• Support sentiment trends. "Support used to be great, now I wait three days" is the canary in the coal mine for a company that's scaling faster than their CS team.
• Unexpected praise. When a competitor's users rave about something you didn't think was a strength, pay attention. Your assumptions about their weaknesses are probably wrong.

Twenty minutes gets you through 5 competitors × 5 reviews each. If you have more competitors, rotate — three this week, two next week. The goal is coverage over time, not completeness every week. For a deeper playbook on catching competitive warning signs in review data, see our guide to CI red flags.

Step 4 — The 20-Minute Changelog & Content Sweep (Monday, 8:45 AM)

Product changelogs and blog posts are your competitor's narrative about themselves. They're not the full truth — changelogs omit the features that shipped broken, the launches that tanked, the strategy pivots that didn't work. But they tell you what your competitor wants the market to believe, which is itself a useful signal.

For each competitor, check:

• Changelog. What shipped this week? Is the velocity picking up or slowing down? Are they shipping core improvements or edge-case polish? Core improvements this week mean they're fighting on product. Polish means the core is stable and they're optimizing. Both are signals — you just need to know which is which.
• Blog. What are they writing about? Blog topics are positioning signals. If a competitor suddenly publishes four posts about enterprise security, they're building an enterprise narrative — probably because they just landed an enterprise deal and want more. If they're writing about AI features they don't have yet, they're pre-announcing a roadmap shift.
• Docs. Skim the docs changelog or "what's new" section. Docs updates precede marketing announcements by weeks. A new API endpoint documented today is a feature launch next month.

Collect links and one-line summaries. Don't write analysis yet. Raw signal collection only. The synthesis comes in Step 5.

Step 5 — The 25-Minute Weekly Brief (Monday, 9:05 AM)

This is the step most teams skip. They collect the signals and then... nothing. The intelligence sits in a doc nobody opens.

The weekly brief is not a report. It's a Slack message or a short email — maximum 400 words. It has exactly two sections:

1. What changed this week (the signals). Bullet points only. "Competitor A dropped Pro plan pricing from $99 to $79/mo." "Competitor B posted 4 ML engineer roles." "Competitor C users complaining about SSO reliability — 3 mentions this week." No interpretation. Just facts.
2. What to do about it (the actions). One sentence per signal. "Sales: Competitor A's new pricing is $20/mo below ours — update battlecard pricing section by EOW." "Product: ML hiring at Competitor B suggests an AI feature launch in Q4 — consider accelerating our AI timeline or preparing a counter-positioning narrative now." "Marketing: Competitor C's SSO complaints are a wedge — add comparison language to competitor landing page."

The format matters. The signal without the action is noise. The action without the signal is unfounded. Together they're intelligence. If a signal doesn't suggest an action — if there's nothing anyone should do differently because of it — it didn't need to be in the brief. Kill it.

Send the brief to the same channel or list every Monday at 9:30 AM. Consistency builds the habit. If your team knows there's a competitive brief in their Slack every Monday morning, they'll start reading it. If it arrives randomly, they won't. Our guide to building a CI program from scratch covers the organizational side in more detail, including how to get buy-in from sales and product teams.

Why 90 minutes is the right number

Ninety minutes is short enough to actually do every week and long enough to catch real signals. If your CI cadence takes four hours, you'll do it twice and stop. If it takes fifteen minutes, you're not going deep enough to find anything useful. Ninety minutes is the sweet spot.

It's also the difference between competitor monitoring and competitive intelligence. Monitoring is collecting raw data. Intelligence is collecting data, filtering it, connecting it to a decision, and delivering it to the person who can act on it. The 90-minute framework forces you through the full intelligence cycle — collect, filter, connect, deliver — instead of stopping at collection and calling it done.

What to measure

A CI program without metrics is a hobby. Track exactly two things:

1. Time-to-detect. How many days between a competitor making a change and your team knowing about it? The goal is under 7 days for pricing changes, under 14 days for hiring surges, under 30 days for product launches. If you're learning about competitor moves from prospects — the classic "oh, did you know Competitor X launched Y?" moment on a sales call — your CI program isn't working.
2. Action rate. What percentage of your weekly briefs produced at least one action — a battlecard update, a roadmap adjustment, a positioning change? If you're sending briefs every week and nothing ever changes because of them, you're producing a newsletter, not intelligence. The action rate is the only metric that measures whether your CI program is useful versus merely active. For the full set of metrics that matter, see the 7 CI metrics every B2B SaaS team should track.

Automation: what to automate, what never to

Steps 1 through 4 are automatable. AI scraping tools can check pricing pages, pull job listings, aggregate reviews, and summarize changelogs without a human in the loop. At RivalSignal, we automate the entire collection pipeline — pricing, hiring, reviews, changelogs — and deliver a pre-processed signal digest. The human only does Step 5: the 25-minute synthesis and action assignment. AI has genuinely changed what's possible in CI.

But Step 5 should never be fully automated. An LLM can draft a brief. It cannot decide what your team should do. That judgment requires context no model has: your product strategy, your sales cycle dynamics, your team's capacity, your board deck commitments. Automate the collection. Keep the decision human.

The ideal CI workflow looks like this: an automated pipeline collects signals across pricing, hiring, reviews, and changelogs on Monday at 6 AM. By 8 AM, a structured digest is in your inbox — raw signals, flagged anomalies, no narrative. You spend 90 minutes on the framework. By 9:30 AM, the brief is in Slack. Your team has competitive intelligence before their second coffee.

The one rule that makes this sustainable

Don't track more than 5 competitors. Don't monitor more than 5 signal types per competitor. The matrix maxes out at 25 cells (5 × 5), and that's already more than most teams can action. If you're tracking 10 competitors across 10 dimensions, you're building a research library, not a CI program. The library will be impressive. Nobody will use it.

Pick your top 3-5 competitors. Pick the 5 signals that actually change decisions at your company — pricing, hiring, reviews, changelog, and one wildcard that's specific to your market. Run the framework every Monday. Skip the weeks where nothing changed — a one-line "No changes this week. Brief resumes next Monday." is better than silence. But if you go three weeks with nothing to report, your signal set is wrong.

Most CI programs die because they try to be comprehensive. This one survives because it's small enough to actually run. Run it for four weeks before you add anything. By then, you'll know what's working and what isn't — and you'll have four weeks of briefs to prove the program's value before anyone asks whether it's worth the time.