How Much Does Your Competitor Sell? 7 Public Signals That Reveal Their Real Size

“How much do they actually make?” is the question every manager has asked while staring at the Instagram of a competitor that looks like it’s booming. It’s also the worst-answered question in the market: either someone sells you a made-up number dressed up as precision, or someone declares that it’s impossible to know. Both answers are wrong.

Let me be blunt: you won’t uncover the exact revenue. That number lives inside the competitor’s ERP and in their tax filings — and nobody is going to hand it to you. But order of magnitude, growth pace and operational health leak from every direction. Brazilian e-commerce moved roughly R$ 235 billion in 2025 (ABComm), and every one of those transactions leaves a public trail. Serious competitive intelligence is the art of reading that trail — not of guessing the round number.

The 7 signals that leak a competitor’s size

1. The order number — the oldest trick and still the best

Make a small purchase today and write down the order number. Make another one seven or ten days later. The gap between the two numbers is, roughly, how many orders the store processed in that window — because most platforms (Nuvemshop, Loja Integrada, Shopify, Magento) generate sequential numbering.

A real example of how the math works: order #018342 on Monday, #019187 ten days later = ~845 orders in 10 days, or about 2,500 orders/month. Multiply by the catalog’s average order value (or by the sector’s R$ 565) and you have a monthly revenue estimate built from two data points any customer can see.

2. Site traffic → sales (the three-factor formula)

SimilarWeb and Semrush estimate the monthly visits of any domain. Take that number and apply the math the entire industry uses:

visits/month × conversion rate × average order value = estimated revenue.

Typical Brazilian e-commerce conversion runs between 1% and 2% (the average sits at the bottom, niche players at the top). So 300,000 visits × 1.5% × R$ 565 ≈ R$ 2.5M/month. The weak spot is conversion: get it wrong, 1.5% versus 1%, and your estimate drops by a third. That’s why it never stands alone — it’s the second vote, not the verdict.

3. Marketplace: the “units sold” and the review count

On Mercado Livre, Amazon and Shopee, the product page usually shows how many units were sold and how many reviews it received. When “units sold” isn’t displayed, here’s the rule of thumb: e-commerce gets 1 review for every 20 to 40 sales — the overwhelming majority of buyers never leave a review.

A product with 1,200 reviews represents, in practice, somewhere between 24,000 and 48,000 units over the listing’s lifetime. Cross it with the creation date and you have the pace — not just the total.

4. Volume of active ads

The Meta Ads Library shows, without logging in, every ad the competitor is running right now (see the step-by-step in how to spy on competitors’ ads). A brand keeping 30 creatives running at once is in a different order of magnitude than one running 3. Paid-media volume is a revenue proxy because nobody sustains heavy media for long without a return that bankrolls it.

5. Social cadence and engagement

Posting frequency, story rhythm, follower growth and — what matters most — absolute engagement (likes and comments per post, not the percentage). A brand that keeps up expensive production and 5 posts a week is allocating a content budget that only adds up above a certain revenue threshold.

6. Reclame Aqui as a transaction counter

Counterintuitive, but it works: complaint volume rises along with sales volume. A store with 400 complaints/month moves far more than one with 12 — even if its reputation is worse. Look at the total and the trend of complaints, not the score.

7. Job postings and the company registry — the signals nobody checks

Open roles on LinkedIn, Gupy and InfoJobs tell you where the money is going: 5 logistics openings = operations expansion; 3 in performance marketing = they’re about to hit the media accelerator. And the company registry (a free lookup at the tax authority or on aggregators) reveals declared size, share capital, partners and founding date — the floor of your estimate.

The golden rule: no single signal is enough

The mistake almost everyone makes is marrying the first number they find. You found the traffic on SimilarWeb, multiplied it, dropped it into the slide. The catch is that this number alone carries the entire error of the conversion rate you guessed.

Where Batedor comes in

The seven signals are public — the problem is that collecting all of them, every week, for each competitor, by hand, doesn’t scale. Batedor automates the part that changes fastest: the public activity (posts, campaigns, frequency, launches) across Instagram, Facebook, YouTube and the website. Instead of opening seven tabs per competitor, the timeline arrives assembled — and whatever signals traction (cadence climbing, a campaign heating up) turns into an alert.

Estimate a competitor in 20 minutes

1. Place a test order and note the number (come back in 7–10 days to close out the orders/month math).
2. Run the domain through SimilarWeb → visits/month.
3. Apply the formula: visits × 1.5% × catalog average order value.
4. Check 2–3 products on the marketplace → reviews × 30.
5. Count the active ads in the Meta Ads Library.
6. See complaints/month on Reclame Aqui to calibrate the volume.
7. Triangulate: throw out the highest and the lowest, keep the middle range.