Strategy & Planning

How to run competitive research as CPG Founders

Strategy & PlanningFor CPG Founders3 apps10 steps~20 min to set up

You're running competitive research the same way you were three years ago: manually checking competitor DTC sites for price changes, screenshotting Amazon listings when you remember to, Googling '[competitor] + deduction dispute' and getting nothing useful, and asking your broker what's moving on shelf. By the time you've compiled anything actionable, a month has passed and the insight is stale. You don't have a market research analyst. You don't have a Nielsen subscription. You have a browser, a Slack channel where you paste links, and a Google Doc that nobody updates. Meanwhile your competitors are iterating on claims, pack formats, and pricing faster than you can track.

Build this on Starch → See the apps used

Strategy & PlanningFor CPG Founders3 apps10 steps~20 min to set up

Outcome

What you'll set up

A browser-automated system that checks competitor DTC sites, Amazon listings, and retail shelf pages on a schedule and surfaces price changes, new SKU launches, and claim updates without you lifting a finger

A weekly competitive digest delivered to your inbox — what changed across your category this week, who's running promotions, and what the X conversation about your competitors looks like

A searchable knowledge base where every competitive finding is automatically categorized, tagged by competitor and insight type, and findable in seconds when you're prepping for a buyer meeting or board update

The Starch recipe

Apps, data, and prompts

The combination of Starch apps, the data sources they pull from, and the prompts you use to drive them.

Apps used

X Mentions Tracker Growth Analyst Knowledge Management

Data sources & config

Starch automates competitor DTC sites and Amazon listing pages through your browser — no API needed. X mentions are tracked daily via browser automation using the X Mentions Tracker app. Notion is connected on a scheduled sync so competitive findings are written back to your team knowledge base and stay searchable. Gmail is connected on a scheduled sync to deliver weekly digests. For any competitor that has a public Shopify storefront, Starch can also query it live from the integration catalog.

Prompts to copy

Every morning, go to [Competitor A]'s Amazon listing and [Competitor B]'s DTC site, record their current prices, any active promotions, and any claim changes on the front of pack. Log everything to a table with today's date.

Every Monday, pull the last 7 days of X mentions for [Competitor A], [Competitor B], and [category keyword]. Summarize sentiment, flag any viral posts, and note any mentions of price complaints, out-of-stock reports, or new product announcements. Email me the summary.

Whenever a new competitive finding is logged, auto-categorize it by competitor, insight type (pricing / claims / distribution / product launch), and confidence level, and add it to our Competitive Intelligence knowledge base in Notion.

Run these in Starch → or paste them into your favorite agent

Walkthrough

Step-by-step

1 List your top 5 competitors by shared retail door count or Amazon subcategory rank — these are your primary targets. Include their DTC URLs, Amazon ASINs, and any known social handles.

2 Install the X Mentions Tracker app from the Starch App Store and configure it to track your competitors' brand handles plus two or three category keywords (e.g., 'better-for-you snack', 'clean protein bar'). It logs daily mentions via browser automation — no X API account needed.

3 Build a browser automation that visits each competitor's Amazon detail page daily, extracts the current Buy Box price, any Lightning Deal or coupon badge, the primary claims on the title and bullet points, and the star rating and review count. Tell Starch: 'Every morning at 7am, go to these five Amazon ASINs, record price, active promo, main claims, and review count, and log to a Google Sheet.'

4 Build a second browser automation for competitors with DTC sites. Tell Starch: 'Visit [Competitor A] and [Competitor B]'s websites weekly, record their hero product prices, any sitewide promo banners, and any new products in their shop nav.' Connect Google Sheets from Starch's integration catalog so findings are logged alongside Amazon data.

5 Set up a Notion database for competitive intelligence using the Knowledge Management app. Tell Starch: 'Create a Notion database called Competitive Intel with fields for Competitor, Date, Insight Type, Source, Raw Finding, and Analyst Note.' Starch syncs your Notion on a schedule so the database stays current.

6 Wire the automations together: after each browser run logs findings to your sheet, trigger a Starch agent to classify each row by insight type (Pricing / Claims / Distribution / New SKU / Social Signal), write the classified record to the Notion database, and flag anything that changed from the prior week.

7 Build a weekly digest automation using Growth Analyst framing. Tell Starch: 'Every Monday at 8am, pull last week's competitive intel rows from Notion, pull last week's X mention summaries, and email me a 300-word brief covering: what changed in pricing, any new SKUs spotted, standout social moments, and one suggested action.' Connect Gmail on a scheduled sync to send the digest.

8 Add a retail shelf check layer if you have a field team. Tell Starch: 'After each field visit, I'll paste photo descriptions into this form. Auto-extract competitor facings, shelf position, and any visible promotions, and log to the Notion database.' This makes qualitative field notes searchable alongside digital data.

9 Before any buyer meeting, ask Starch to pull everything in the Notion Competitive Intel database tagged to that retailer's channel (e.g., 'natural grocery') and generate a one-page summary of what your top three competitors have done in that channel over the last 90 days.

10 Before board meetings or investor updates, use Presentation Agent (currently in development — request beta access) to turn your Notion competitive summary into a clean slide. In the meantime, tell Starch: 'Pull the last 30 days of competitive intel from Notion and write a 5-bullet exec summary I can paste into slides, including any pricing moves, new entrants, and category trends from X.'

See this running on Starch

Connect your tools, describe what you want, and the agent builds it. Closed beta is free.

Try it on Starch →

Worked example

Q1 2026 category review — better-for-you snack bar brand

Sample numbers from a real run

Competitor A — Amazon price drop detected	2
Competitor B — new 'no added sugar' claim on front of pack	1
Competitor C — out-of-stock signal from X mentions	14
New entrant spotted in DTC shelf scan	1
Category keyword mentions on X (week of March 3)	847

During the week of March 3, the Monday digest flagged that Competitor A had dropped their 12-count Amazon price from $28.99 to $24.99 — a 14% cut that put them below your everyday price for the first time. Simultaneously, the browser automation caught Competitor B quietly swapping their hero claim from 'high protein' to 'no added sugar' on both their DTC homepage and their Amazon title. The X Mentions Tracker logged 14 tweets in four days complaining that Competitor C was out of stock at Target — an opening in natural grocery your broker hadn't surfaced yet. All three findings were auto-classified, written to the Notion Competitive Intel database, and included in the Monday digest before your weekly team standup. You walked in with a pricing response recommendation, a claims audit question for your packaging team, and a lead for a Target velocity conversation — none of which required a single manual Google search.

Measurement

How you'll know it's working

Competitor price index — your shelf price vs. top 3 competitors across Amazon and DTC, tracked weekly

Claim drift rate — how often competitor front-of-pack or Amazon title claims change quarter over quarter

Category share of voice on X — your brand mentions vs. competitors' over rolling 30 days

New SKU launch lag — days between a competitor's new product appearing in your scan and your team being aware of it

Retail out-of-stock signals — competitor OOS mentions from X or field scans, which indicate a distribution opportunity

Comparison

What this replaces

The other ways teams handle this today, and how the Starch version compares.

Mintel / Nielsen / SPINS subscription

Deep category data but starts at $20k+/year, lags by weeks, and doesn't give you real-time DTC or Amazon signal — built for brand managers with research teams, not a founder doing this between co-packer calls.

Manual spreadsheet + browser bookmarks

Zero cost but only gets checked when you remember, misses subtle claim changes, and lives in no one's head consistently — the insight rot is the real cost.

Crayon or Klue (competitive intelligence SaaS)

Good for software companies tracking feature pages, but limited coverage of Amazon listings, DTC storefronts, and physical retail signals that are the actual competitive surface for a CPG brand.

Brand24 or Mention for social listening

Handles social mention volume well but doesn't connect to your pricing data, Notion knowledge base, or retail scan findings — you'd still be stitching outputs together manually.

On Starch RECOMMENDED

One platform — x mentions tracker, growth analyst, knowledge management all running on connected data. Setup in plain English; numbers stay current via scheduled syncs and live agent queries.

Try it on Starch →

FAQ

Frequently asked questions

Can Starch actually monitor Amazon listings without an Amazon API key?

Yes. Starch automates Amazon product pages through your browser — no API needed. It navigates to the ASIN, reads the Buy Box price, badge state, claims, and review count the same way you would manually, and logs it. The one thing browser automation can't give you is Amazon's internal sales rank history or ad spend data — for that you'd need a tool with Amazon SP-API access, which Starch doesn't currently offer.

What happens if a competitor's DTC site has a bot blocker?

Some sites block automated browsers more aggressively than others. For heavily protected sites, Starch may return incomplete data or fail gracefully. In practice, most Shopify-based DTC storefronts — which covers the majority of CPG brands — are reachable without issue. If a specific site is blocked, you can flag it and Starch will try alternate scraping approaches or surface that data gap clearly rather than silently failing.

Is Starch SOC 2 certified? My co-packer requires vendor security documentation.

Not yet — Starch is not currently SOC 2 Type II certified. If your co-packer or a retail partner requires SOC 2 documentation for a vendor handling sensitive operational data, that's worth knowing upfront. For competitive research specifically, the data flowing through is largely public-facing competitor information, which is lower sensitivity than, say, financial or employee data.

Can I track competitors on TikTok Shop or Instagram, not just X?

X Mentions Tracker is the pre-built app for social monitoring. For TikTok Shop or Instagram, there's no pre-built template today, but Starch can automate public TikTok and Instagram pages through your browser — no API needed. You'd describe what you want ('check this TikTok creator's latest posts weekly and log any brand mentions or new product tags') and Starch builds it. Coverage depends on what's publicly visible without a login.

Will this replace a SPINS or Nielsen subscription for category data?

No — and it's worth being direct about that. SPINS and Nielsen give you actual point-of-sale velocity data across thousands of retail doors, which no browser automation can replicate. What Starch gives you is the real-time digital signal layer: pricing moves, claim changes, social conversation, and DTC activity. For a Series A or earlier CPG brand, that's often more actionable week-to-week than scan data that's 8 weeks old by the time you see it. The two are complementary once you can afford both.

How do I get competitive findings into a format I can share with my broker or board?

Everything logged to your Notion Competitive Intel database is searchable and exportable. For broker updates, tell Starch to pull the last 30 days of findings for a specific retail channel and write a summary. For board decks, Presentation Agent — currently in development, with beta access available — will turn a Notion summary into slides automatically. Until then, Starch can write a structured brief you paste directly into your existing deck template.