Knowing when a competitor is about to launch a product sounds like spy work, but in practice it is trail reading. No e-commerce launch comes out of nowhere: weeks before D-day there is already a teaser on Instagram, a new page on the site, a new menu category, a creator receiving a press kit. Whoever tracks these signals discovers the launch a week or two in advance. Whoever does not finds out in the feed, along with the rest of the market, when the only possible response is to improvise.
This guide shows the trails that precede a launch, how automatic detection catches each of them and, the part almost nobody writes about, what to do with the two-week head start you gain.
Every launch leaks (and it leaks in a predictable order)
A launch is not a post: it is an operation. The product has to be registered in the store before it can receive paid traffic, the page has to be indexed before search works, the creator has to receive the product before recording the video. Each of these operational needs leaves a public trail, and the order of those trails repeats from brand to brand, because the logistics are the same for everyone.
The typical trail of an e-commerce launch
D-30 to D-21
Creators receive the product
Press kits and “gifted product” posts show up in stories tagging the brand, before any official announcement. In beauty and fashion, it is almost always the first visible signal.
D-21 to D-14
The page goes live with no promotion
Product registered, a new URL in the sitemap, sometimes a “coming soon” landing page. The store needs this ready before running ads.
D-14 to D-7
The site menu changes
A new category, a reorganized collection, a reserved secondary banner. A change in menu architecture rarely happens without a reason.
D-7 to D-1
Public teaser
A countdown in stories, a “coming soon”, an email or WhatsApp waitlist.
D-day
Coordinated launch
A feed post, an email to the list, the main homepage banner, creators posting in sequence.
Fonte: Synthesis of common online retail practices; the timeline varies by brand and category
The timeline compresses or stretches depending on the size of the brand: a small operation handles everything in ten days, a large one stretches it past a month. But the sequence holds, and that is what matters. If you have seen the page go live and the menu change, the teaser is a matter of days. In creator-driven segments, as we show in the guide to competitive intelligence for beauty e-commerce, the cycle of gifted product posts can hand you the launch a month ahead, because the physical product has to travel to the influencer.
How to detect a competitor’s launch automatically
You can watch all of this by hand: check the sitemap every Monday, open each competitor’s stories every day, compare the menu from memory. The manual method, with sitemap.xml and the Wayback Machine, is detailed in the guide to how to monitor a competitor’s website. It works, and it is free. The problem is consistency: the trail appears on Tuesday and you check on Saturday. The alternative is to let a bot run the patrol and save your time for deciding what to do with the finding.
In Batedor, the scan runs 24/7 over the public content of each competitor’s Instagram, Facebook, YouTube and website, and the AI classifies every detection into 16 types: promotion, coupon, free shipping, stock clearance and, what matters here, “launch”. On the campaign timeline, the story teaser and the new site banner arrive already classified, filterable by type. That alone solves the Saturday problem. But the most interesting signals come from the analysis layers on top.
The new-type signal (the most valuable one)
On the panel’s Strategic signals page, statistical detectors run every night over up to 90 days of each competitor’s campaigns. One of them looks for exactly the pre-launch pattern: a campaign type that has never appeared in the historical base. If a competitor that only ran coupons and free shipping starts publishing campaigns classified as launches, and that repeats across two or more detections within 14 days, an anomaly signal is created with the evidence attached; at three occurrences the severity rises to medium, at five, to high.
Pivot and volume: the warm-up shows in the numbers
Two other detectors round out the picture. The pivot detector compares the distribution of campaign types over the last 14 days with the 90-day baseline: a divergence above 25% becomes a signal (“the mix shifted from discount to launch”), with severity rising along with the size of the change. And the volume anomaly flags when the current week strays from the historical average by more than two standard deviations, up or down. A launch warm-up usually lights up both: the competitor posts more than usual and about a topic that is not the usual one.
Strategic move: the read of the week
Finally, once a day the AI compares today’s dossier with the one from seven days ago and summarizes the most significant change on the Moves page: a change in tone, in cadence, a new channel or a change in format (the classic case: the campaigns stopped being promotional and turned into product teasers). When the significance is medium or high, it becomes a real-time notification in the panel’s bell, on top of the email alerts and the Slack, Telegram and WhatsApp channels. In seasonal mode, this check starts running hourly. You can see this working with your real competitors in the 14-day trial, no card required: the first detection shows up in the panel within minutes, and the statistical signals gain precision as the history builds up.
What automatic detection does not catch
Honesty before the playbook: none of this is a crystal ball, and the limits are real.
- Public content only. Internal meetings, DMs, closed creator groups and agency briefings leave no monitorable trail, and they should not: the collection respects the LGPD (Brazil’s data protection law) and is limited to what the competitor publishes openly.
- It needs history. The statistical detectors compare the recent period with the baseline: without a few dozen campaigns accumulated over 90 days, the new-type signal and the pivot detector still have no basis for comparison. The move detection needs at least seven days of dossier.
- A vague teaser confuses the classifier. A “coming soon” with no mention of a product may land as brand content until the competitor gives more context. The classification improves as the launch approaches and the trail becomes explicit.
- A silent competitor generates no signal. Anyone who launches with no teaser, no creator and the page going live the day before (rare, because it wastes the launch itself) will only be detected on D-day.
It is also worth noting that strategic signals and move detection are not on every plan; the full grid is on plans.
You have two weeks: the response playbook
The advantage of detecting early is not the gossip, it is the reaction time. Two weeks are enough to do a lot, as long as you do not spend ten of them in meetings. The most common mistake is the opposite of panic: knowing and doing nothing, because “it is not official yet”. A page registered on the competitor’s site is as official as the information will get before D-day.
| When | Action | Why |
|---|---|---|
| Days 1 and 2 | Confirm and size it up. Is it a new line, an extension or a repositioning? What price range does the registered page suggest? | Reacting to a rumor burns energy; reacting to a live page is strategy. |
| Days 3 to 5 | Decide the pricing stance for your equivalent product: defend, hold or differentiate, with the margin calculated. | A pricing decision made calmly costs less margin than one made in the shock of D-day. |
| Days 6 to 9 | Build the counter-offer: a kit, a bundle or a benefit on your best-seller in the same category. | You will be competing for the same buyer’s attention in the same week. |
| Days 10 to 13 | Schedule your own content for the launch week and secure stock of your equivalent. | An empty feed during your rival’s launch week is free share of voice for them. |
| D-day | Do not touch the price. Monitor the reception: engagement, comments, stockouts. | The first week tells you whether the launch caught on; reacting before that is a guess. |
A concrete example from the Brazilian calendar: if a skincare competitor is going to launch a hair care line aimed at Christmas, the page has to be live in November, the creators receive the product in October and the new-type signal tends to light up weeks before the first official post. Whoever monitors gains all of November to prepare the response kit; whoever does not finds out in December, with the stock already bought and the content calendar locked.
See your first competitor in minutes
14-day free trial, no card. Within minutes, the first detection shows up on your dashboard.
Create free account