From 77% to 100% coverage: the consumer goods playbook for selling 24% more with AI agents and without hiring more sales reps

By Alejandro Pascual ("Unfiltered AI" newsletter, LinkedIn) · Mon, 12 Oct 2026.

The article argues that a consumer goods company currently covering 77% of its market can sell 24% more without hiring a single sales rep, and that one with 60% coverage could sell half as much again. According to the author, this is not a consultant's promise but simple multiplication. Beyond sales, he promises higher margins (by no longer visiting points of sale that lose money and no longer granting discounts nobody asked for) and greater penetration (by not ceding shelf space to competitors).

The author proposes five moves. The first is to take a census of the market rather than of the customer list: map every point of sale in the territory using satellite and street-level imagery, with a model that detects, locates and classifies stores in weeks rather than months. The resulting real coverage is almost always lower than people think.

The second is to decide whom to serve. Each outlet gets a score (how much it could buy in the category minus the cost of serving it) and is ranked. Where the contribution stops growing, you stop. According to the author, the last 10% is almost never profitable and is left for step five.

The third is to assign the channel based on buying patterns. A visit costs 5 to 15 times more than a remote contact. Large accounts get visits; mid-sized ones, visits plus remote contact; small ones, remote contact; tiny ones, an agent. With AI, frequency is decided every week rather than every year.

The fourth is to grow the order, not just the visit. An order is suggested based on purchase history, the missing category and the active promotion, and the sales rep confirms it. That, the author says, is where real penetration is won.

The fifth is to serve the long tail with an agent. Outlets that do not justify a visit do justify WhatsApp or messaging: the agent takes the order, suggests the mix and reminds them when to restock. The marginal cost is close to zero, which lowers the cost to serve and raises the ceiling of step two.

As a bonus for those seeking the full margin, he proposes governing price: drawing the waterfall from list price to pocket price (invoice discount, rebate, early payment, absorbed freight, credit). He cites the example of a sales rep who "gave 6%" when the customer actually received 15%. With a 12% margin, a 4% price cut requires 50% more volume to break even, something that almost never happens. The solution is a price corridor with a floor and approval.

Where the figure comes from: according to the author, in consumer goods each point of distribution moves between 0.6 and 1.0 points of volume, with 0.8 as the working value. Going from 77% to 100% means 30% more distribution; multiplied by 0.8, that yields 24% more sales. On penetration, he cites the largest study of stockouts (71,000 consumers in 29 countries), with an average stockout rate of 8.3%: when the product is not there, 26% buy another brand and 31% go to another store. He also mentions a point-of-sale prioritization program in Latin America in which sales rose 40% and penetration 25 points.

As reference cases he cites AB InBev, which created BEES, its B2B platform for neighborhood stores. According to the text, replenishment orders moved to mobile, sales reps stopped taking orders and started selling mix, 75% of its revenue now flows through digital platforms, and BEES handled $49 billion in 2024. He also cites Coca-Cola FEMSA with Juntos+, which has 1.3 million monthly active users and an AI advisor already working with 70% of the sales force, telling each rep which outlet to visit and what to order. And Fastenal, in industrial distribution, which closed branches and grew: large customers moved to Onsite locations within the customer's premises and recurring consumption shifted to vending machines.

The author recounts a case of his own with the sales director of a consumer goods company in Central America. His presentation said 30% coverage; when the census was done with street imagery, 11,000 points of sale turned up that were not in any database, and real coverage was 18%. They deployed a WhatsApp or messaging agent (catalog, ordering, reminders) for the 4,000 smallest outlets, without hiring new sales reps. After 90 days, 1,900 of them had bought at least once, the cost per order was 14 times lower than that of a visit, and the field force, freed from the small accounts, increased the average order among large accounts by 9%. All these figures are the author's own claims.

Finally, he describes six components for implementation, in this order: census and diagnosis (universe, real coverage and contribution per outlet, three figures in 90 days); route design (segmentation by potential and cost to serve, territories drawn by potential rather than by map); focused field force (reps keep A and B accounts, visit fewer outlets and sell more mix); a commercial contact center with assigned portfolios and three teams (farmers for the active base, hunters for new accounts in the white space revealed by the census, and reactivation of accounts that stopped buying); a digital agent for the long tail, which the author says is the piece that raises the ceiling; and a control tower with one definition per metric, alerts with an owner, and a cadence (daily for exceptions, weekly for performance). On top of that, optionally, the price corridor and discount governance.

The piece closes by inviting readers to contact the author with the question: "Is your 30% real, or is it the number in the presentation?" It is an opinion piece with a commercial tone, and the figures in the email have not been independently verified.