Competitive Intelligence for Consumer Goods That Predicts Share

Competitive Intelligence for Consumer Goods That Predicts Share

The Australian sauces brand had been running a clean dashboard for eighteen months. Brandwatch tracked sentiment across the category. Meltwater measured share-of-voice against the two largest competitors. Google Alerts pinged the marketing team whenever retailer trade press mentioned the brand by name. The CMO ran a weekly intelligence review every Monday at 9am, and not once in that whole period did the dashboard suggest the company was about to lose two share points in a single quarter.

In April, a finance team member pulled the latest Circana scan-data report and stopped breathing. Volume share across Coles and Woolworths had dropped from 8.4% to 6.4% over the prior quarter. Margin had held, but only because the brand had quietly stopped funding its trade-marketing reserve to protect the P&L. Three months had already passed. The board meeting was in eleven days.

The actual cause was sitting in plain view inside the retailer apps the CMO had not opened in months. A private-label sauces competitor had picked up sixty additional Coles stores during the February range review, expanding its facing count by roughly 40%. None of those decisions were announced on social media. None of them generated trade-press coverage. The share-of-voice dashboard kept showing the brand as the category leader, because the brand still posted the most content. Content, by that point, did not matter. The shelf had decided.

This story is composite, but the pattern is not. I have watched four Australian consumer-goods brands lose share between 2023 and 2025 the same way. Each time, the CMO and the trade-marketing director were tracking the wrong signals. Each time, they discovered the loss three months too late. And each time, a competitive intelligence for consumer goods program built around social listening was the proximate cause of the blindness.

The cost of the lag is not just the lost share points. It is the cascade that follows. When you find out late, you negotiate from weakness. The buyer at Coles or Woolworths already knows your facings are at risk and has already had three trade conversations with the rival. Your trade-marketing dollars get redeployed in a panic, often into deeper promotions that erode reference price for the next eighteen months. The brand walks out of the next range review with a smaller footprint and a thinner margin, both consequences of an intelligence system that was looking at the wrong screen.

Why the Math Doesn't Work: Social Listening Cannot See Shelves

Social-listening platforms were built in the text era, for marketing teams who cared about sentiment, brand chatter, and PR risk. They were never engineered to detect shelf decisions, range changes, or promotional depth. A 2025 platform comparison from Brandwatch vs Meltwater describes the underlying architecture as text-first and acknowledges that video and physical-shelf signals fall outside what the systems were designed to capture. The 2025 Britopian listening review of consumer-intelligence platforms makes the same observation: the gaps for CPG categories are structural, not configurable. You cannot dashboard your way out of them.

That mismatch shows up in the macro numbers. Circana data for Australia, summarised in a Private label expansion feature, found that private-label brands represented 36% of Australia's CPG/FMCG sales at the end of 2024. The category had grown by 4.8 percentage points year-on-year, worth roughly $46 billion in sales. None of that share movement was driven by social media activity. It was driven by retailer-controlled distribution, shelf placement, and base-price decisions across Coles, Woolworths, and ALDI. A brand watching Meltwater alerts during 2024 would have seen its absolute mention count rise while its volume share fell, because the rival was a retailer-owned label that does very little above-the-line communication.

The mistake compounds when budget meets blindness. Most consumer-goods brands inside the $1M-$10M revenue band cannot afford a Nielsen scan-data subscription, so they treat their social-listening tool as the proxy. The platforms themselves know what they cannot see. Their product roadmaps confirm it: most of the major listening tools are pivoting toward video and decision-stage signals because their text-only roots have run out of road. The problem is that the operators using them generally do not.

Reactive Coles Circana data commentary on Australian retailer share movements reinforces the same point. The levers that move share live inside shelf and promotional decisions, not inside chatter volume. When the social dashboard says everything is fine and the scan data says you have lost two points, the scan data is right. By the time it tells you, you have already paid for the loss. That is the lie at the centre of most consumer-goods CI programs in 2026.

The Shelf Signal Intelligence System Blueprint

I started using The Shelf Signal Intelligence System with brands that could not justify a Nielsen subscription but could not afford to keep being surprised by share decline. The system tracks four signals that predict share shift before it lands on the P&L: distribution gains, base-price moves, promo depth, and new-SKU velocity. Each signal has a measurable weekly cadence. Each signal can be captured manually, semi-automatically, or via a paid tool, depending on the budget the brand can deploy.

Distribution gains are the strongest single predictor. A competitor that adds twenty stores in a range review will outpace any brand that depends on the lost facings within four to eight weeks. The signal sits at the store-count level, not the SKU-velocity level. You can capture it through scheduled store walks, retailer-app monitoring, or platforms like Stackline platform that aggregate digital-shelf SKU presence across retailers. Stackline's Stackline Atlas product page documents weekly market-share intelligence across more than a billion products, which is overkill for a $5M Australian sauces brand but useful as a benchmark for what enterprise CI looks like.

Base-price moves are the second signal. A 50-cent reduction on a competitor's hero SKU, held for three weeks, will pull volume across the category and reset the consumer's reference price for the next twelve months. Most $1M-$10M brands track promotional pricing but ignore base-price drift. The two are different. Promotion runs end. Base-price moves stick.

Promo depth is the third signal. A 25% promotion run twice a year is a margin-friendly sales lift. A 25% promotion run six times in nine months tells you the competitor is defending against a private-label entry or trying to disqualify your SKU from the next range review. The signal is not the depth itself, but the change in depth-and-frequency over time, watched across a rolling twelve-week window.

New-SKU velocity is the fourth. The cadence of new-product launches inside a category, especially private-label launches, predicts where shelf reallocation is headed. A retailer that approves three new private-label SKUs in a category over six months is preparing to displace at least one branded SKU during the next range review. Bowen private label describes how Coles and Woolworths run their private-label range cycles, and the velocity pattern is consistent across categories.

The four signals work as a portfolio. No single one is causal on its own. Together they build a leading indicator that lands six to twelve weeks before the share number lands. Pygmalios reports in its Pygmalios shelf AI review that 99% precision is achievable on shelf monitoring via AI, with around 20% of FMCG store audits in 2025 conducted via AI, robotics, or drones. That number will keep climbing. The brands that wait for it to become standard before adopting any version of shelf-level CI are the brands that will keep getting surprised.

A note on weighting. In the categories I have run this in, distribution gains carry the most predictive weight, followed by base-price moves, then promo depth, then new-SKU velocity. The exact ratio shifts by category. In ambient grocery (sauces, condiments, spreads), distribution and base-price dominate. In chilled (yoghurts, dips, cheese), promo depth carries more weight because consumers stockpile less and respond more to in-week price moves. In personal care, new-SKU velocity rises in importance because the category is more fashion-driven. Your team needs to recalibrate the weighting once a year, anchored against the quarter where you saw the largest share movement, working backwards to ask which signal moved first.

Execution: Day 0 to Day 90

The build runs in three phases. Phase 1 is manual and free. Phase 2 introduces paid data layers. Phase 3 reorders the brand's intelligence priorities so that social listening sits where it belongs, which is well below the four shelf signals.

Phase 1: Days 1-30. Build the manual signal layer.

Pick four flagship metro stores per retailer. Two Coles, two Woolworths, both inside the categories where you compete. Assign one team member, ideally the trade-marketing manager or a senior brand analyst, to walk the same four stores every Wednesday morning. The walk takes ninety minutes total per retailer. They photograph the planogram, log facings count by SKU, log shelf price, and log any promotional collateral. The output is a two-page weekly report, attached to a shared folder with timestamped photos.

Run a price-and-promo scrape script against the Coles and Woolworths online stores. The two retailers expose enough product data through their public web pages to support a twice-daily pull on every SKU in your category and your top three competitors. The walkthrough at Scraping Coles Woolworths covers the AU-specific approach. A junior analyst with basic Python or a contractor on a one-month engagement can stand it up. Budget: $1,500-$3,000 once-off, then negligible to run.

Build a single Google Sheet with one tab per signal. Distribution gains tracks store count by SKU, week over week. Base-price moves tracks shelf price ex-promo by SKU. Promo depth tracks discount percentage and weeks-on-promo. New-SKU velocity tracks new launches in the category by week. Conditional formatting handles the alerting. Anyone in the team can read it in five minutes. The Shelf Signal Intelligence System dashboard does not need a BI tool. A Google Sheet with named owners and weekly check-ins beats a Tableau license that nobody updates.

By day thirty, you have a four-signal dashboard, a weekly cadence, and a named owner. You will already have caught at least one pricing or distribution move you would have missed at week zero. That early catch is the proof point you take to the leadership team to fund Phase 2.

Phase 2: Months 2-3. Layer in scan data and digital-shelf intelligence.

Add a partial scan-data feed. Most $1M-$10M brands cannot justify a full Circana or NielsenIQ subscription, but partial feeds, category subsets, or shared-cost subscriptions through brokers are accessible at $15,000-$40,000 per year. Use the scan data to validate the manual signals. If your store walks suggest a competitor has gained distribution, scan data should confirm it within two to three weeks. If the two layers disagree, trust the store walk first and investigate the scan-data lag before reacting.

Add a digital-shelf intelligence layer for online-channel signal capture. Single-seat platforms in the Stackline category run $1,000-$3,000 per month and capture velocity, price, and ranking data on retailer e-commerce sites. The Australia retailers AI profile on the Coles, Woolworths, and Wesfarmers retail-media power-trio describes how AI is reshaping retailer decision-making, which means digital-shelf signals are becoming earlier indicators of physical-shelf moves than they were even two years ago.

By the end of month three, the four-signal model has two reinforcing data layers. Manual store-walk data tells you what is happening this week. Scan and digital-shelf data tells you whether the move is national or regional, durable or temporary, and whether it is showing up in the duopoly's own internal performance numbers.

Phase 3: Quarter 2 onward. Demote social listening.

Cancel the Brandwatch or Meltwater contract or move it down to the cheapest tier. Reassign the budget into the scan-data and digital-shelf layers. Tell the marketing team that social listening now feeds creative briefs and PR-risk monitoring, full stop. It does not feed competitive intelligence. The four shelf signals do.

The cultural reset matters as much as the budget reset. Most consumer-goods brands inherited the social-listening dashboard from a 2017-era marketing roadmap that never got rebuilt. By formally moving it out of the CI category, you stop the false-confidence problem at its source and free up roughly $24,000-$60,000 a year that should have been funding shelf-level signal capture in the first place.

From Quarterly Surprise to Weekly Share Defence

The old state: a brand that discovered share decline three months after it hit the P&L, after the trade-marketing budget had already been cut, after the board meeting had already been booked. Recovery in that state takes two quarters minimum, and the brand often pays for the lost facings with a permanent margin concession the buyer will never give back.

The new state: a brand that catches a competitor distribution gain in the week it happens, escalates within forty-eight hours, and has a counter-promotion or buyer-conversation drafted before the next planogram review. Recovery in that state takes weeks, not quarters, because the move is contested while it is still small.

The metric that defines the new state is leading-indicator share movement. You stop asking "what was our share last quarter" and start asking "what is our four-signal index this week, and what does that predict for share twelve weeks from now." The first question is a postmortem. The second is a defence plan. The Shelf Signal Intelligence System exists to push your team from the first question to the second on a permanent weekly cadence.

The brands that hold share in Australian consumer-goods categories through the rest of this decade will be the ones who treat shelf signals as their primary competitive radar and treat social listening as a creative input. The math has already decided this. The question is whether your team builds the four-signal layer this quarter, or finds out about the next 60-store private-label move three months after it ships.