Skip to content
Product
PlatformPricingDimensionsIntegrationsSecurityCompare
Solutions · by use case
Customer, guest and fan experienceProduct feedback and prioritizationRevenue, churn and retentionAgenciesContact center performanceAI agent and chatbot analyticsMulti-location performanceReviews and competitive benchmarking
Solutions · by industry
Hotels and hospitalityRestaurants and food serviceFitness and leisureRetail and e-commerceSports and live entertainmentInsuranceBanking and financial servicesSoftware and SaaSAutomotiveConsumer products and CPGContact centers and BPOsPublic sector and nonprofitReal estateGaming and betting
Learn
Artifacts GalleryCustomer storiesResearchBlogThought leadershipLessons Log in Start now
Blog

Changing a Recipe Customers Already Buy: How Do They React?

The loudest backlash to a recipe change does not go with the worst sales. Four dated cases, three studies on whether people notice, and how to test a reformulation in a few markets first.

Customers react to a recipe change in two ways that can point in opposite directions: some complain, and some quietly stop buying, and the complaints are the weaker guide to the sales. Coca-Cola's 1985 formula change drew so many calls that the original was back in 79 days, while Diet Pepsi's volume fell for the year after its 2015 sweetener switch, a change PepsiCo said customers wanted. For a food, drink or restaurant brand planning a reformulation, Dimension Labs reads every care contact, review and survey comment about the change, separates customers who tasted a difference from customers who heard about one, and joins both groups to what they bought next.

This page uses four dated public cases, read at the company's own account or in reporting from the time, and three studies of whether people notice gradual changes. Taking a product off the menu altogether is covered in launching or cutting a product; making portions smaller instead of raising prices is in smaller portions or higher prices.

What happens when a company changes a recipe customers already buy?

The four best-documented cases split two ways: on whether customers objected loudly, and on whether sales held. They did not line up.

Dated public cases

What customers said did not predict what sales did

Four reformulations of products customers already bought, what customers said, and what the sales record shows.

ChangeWhat customers saidWhat the sales record showsWhat the company did
Coca-Cola's new formula, April 23, 1985Hotline calls up from 400 to 1,500 a day; protest groups; hoardingNo sales figure in the company's accountOriginal returned as Coca-Cola classic after 79 days
Diet Pepsi drops aspartame, August 2015PepsiCo said customers wanted aspartame-free drinksVolume down 5.8% in 2015 and about 11% in Q1 2016 (Beverage Digest)Aspartame version announced as a second product, June 2016
Kraft Mac & Cheese drops artificial dyes, December 2015Very few people noticedOver 50 million boxes of the new recipe sold before the revealKept the recipe and announced it afterwards
Irn-Bru halves its sugar, January 2018Petition with more than 8,000 supportersRegular Irn-Bru volume share of UK carbonates up 4.2% after the changeKept the recipe
Sources: Coca-Cola history; AFP, June 27, 2016; Fortune, March 7, 2016; Reuters, January 4, 2018; A.G. Barr results, March 26, 2019. All read November 20, 2026.

Coca-Cola announced a new formula on April 23, 1985, its first change in 99 years, and brought back the original as Coca-Cola classic on July 11, after 79 days of protest. PepsiCo said in April 2015 that Diet Pepsi would switch from aspartame to a sucralose blend that August; Beverage Digest data cited at the time showed Diet Pepsi's volume down 5.8% in 2015 and about 11% at retail in the first quarter of 2016, and in June 2016 PepsiCo announced an aspartame version as a separate product. Kraft's macaroni and cheese replaced artificial dyes with paprika and turmeric from December 2015 and sold more than 50 million boxes before the company said anything on the box or in its advertising. And Irn-Bru, the Scottish soft drink, cut its sugar from 10 grams per 100 ml to just under 5 in January 2018, drawing a petition of more than 8,000 people. A year later its maker reported that regular Irn-Bru had gained 4.2% in volume share of the UK carbonates market since the change.

So the amount of complaint did not predict the sales result. Irn-Bru drew a petition and kept growing; Diet Pepsi made a change its maker said customers wanted, and lost sales. What separated them, as the sections below show, is who noticed the change in the product itself, and whether they had somewhere else to go.

Why did New Coke fail when taste tests favored it?

The taste tests measured the wrong thing. Coca-Cola's new formula was preferred in taste tests of nearly 200,000 consumers, but, in the company's own words, those tests did not show the bond customers felt with the original, which they did not want anyone changing. The company had a reason to try: its share lead over its main competitor had been slipping for 15 straight years.

The reaction came through the channel a modern brand would call customer care. By June 1985 the consumer hotline was taking 1,500 calls a day, against 400 before the change, and in the two days after the return was announced it took 31,600. That is 3.75 times the normal load within weeks, and about 40 times it on the days of the reversal.

From Coca-Cola's own account, counted by Dimension Labs

Calls rose almost fourfold after the change and about fortyfold after the reversal

Calls a day to Coca-Cola's consumer hotline in 1985. The last bar averages the 31,600 calls received in the two days after the original formula's return was announced.

Before the change
400
By June 1985
1,500
After the return was announcedJuly 1985, two-day average
15,800
Source: Coca-Cola history, read November 20, 2026. Daily averages and ratios computed by Dimension Labs from the page's figures.

The calls were both a warning and a measure of attachment: customers called to complain, and called in far greater numbers when the original came back. Coca-Cola's account publishes no sales figures for those 79 days, so the record shows the strength of feeling and the decision, not the size of any loss. A brand changing a recipe today has much more than a call count. Every contact, review and comment can be labelled for what the customer is reacting to, and joined to whether that customer kept buying.

Does a backlash mean sales will fall?

Not on its own. Irn-Bru's petition was real, yet its maker's results for the year after the change show revenue up 5.6% to £279.0 million and regular Irn-Bru gaining volume share. Two things help explain it. Most buyers carried on buying, as the share figure shows, and the change happened as the UK's new levy on sugary drinks was taking effect, which put 99% of the company's soft drinks outside the tax. A sales result after a reformulation always comes mixed with what else changed in the market that year, which is why it has to be read against a comparison; was it the decision or the weather shows how.

Diet Pepsi is the opposite pattern. The switch away from aspartame answered customers who said they wanted it gone, yet Diet Pepsi's volume kept falling, at a time when diet colas as a whole were struggling, and the company's statement on bringing aspartame back was that "Consumers want choice in diet colas". When the customers who ask for a change are not the customers who buy the product, a survey of stated preference can point the wrong way. Do customers do what they say? covers how to check stated intentions against purchases.

The useful split is between customers who noticed a difference in the product and customers who object to the idea of the change. The first group shows up in comments about taste, texture, sweetness or how a dish "isn't the same", and in repeat purchase. The second shows up in comments about the announcement, often from people who had not tried the new version. Both deserve a reply; the first is the group that shows up in the sales line.

Should you announce a recipe change or make it quietly?

The evidence favors changing first and explaining after, as long as the change is small enough not to be noticed. Kraft told the market in April 2015 that the change was coming but put no new label or advertising on the product until more than 50 million boxes had sold; its vice president of meals said the company wanted fans to try the new recipe "without being prompted".

Published research, read at the source

Small, gradual changes went unnoticed; telling people a food is healthier can lower how much they enjoy it

Three studies that bear on how and whether to tell customers about a recipe change.

StudyDesignWhat it found
Girgis et al. 2003Randomized trial, 110 people, bread sodium cut 5% a week for six weeksNo more likely than the control group to notice a difference; no change in liking
El Ati et al. 2021Pilot in one Tunisian city over three yearsSalt in bread cut 35% without consumers detecting it
Raghunathan, Naylor and Hoyer 2006Four experimentsThe less healthy a food is portrayed to be, the better people expect it to taste and the more they enjoy eating it
Abstracts read through PubMed and the publishers' Crossref records, November 20, 2026.

Two lines of research explain why that worked. Gradual changes slip under the threshold of notice: in a randomized trial, people eating bread whose sodium was cut 5% a week for six weeks were no more likely than a control group to report a difference, and liked it as much; a three-year program in one Tunisian city cut salt in bread by 35% without consumers detecting it. And labels change the experience: across four experiments, the less healthy a food was portrayed to be, the better people expected it to taste and the more they enjoyed eating it. Announcing "now with less sugar" can make customers taste a loss that a blind tasting would not find.

Ingredient lists still change, so a brand needs an answer ready for the customers who ask. Irn-Bru cut sugar by about 50% at once, and calories by 52.9%, a change large enough to be noticed whatever the company said, which is why its reaction was loud.

How do you test a recipe change before it reaches every store?

  1. Taste-test blind and branded. Coca-Cola's tests measured taste; the attachment showed once customers knew the formula had changed. Run both kinds, and treat a gap between them as the size of the risk.
  2. Change a few markets or stores first. Keep the old recipe in comparable markets over the same weeks, so the season and the economy fall on both sides. How to pick pilot locations covers the choice.
  3. Make large changes in steps. The bread trial cut 5% at a time. A cut taken in steps can be paused if one step draws comments about taste.
  4. Label every contact about the product. Read care calls, chats, emails, reviews and survey comments every week, and keep "tastes different" apart from "heard you changed it". Count each group by market, pilot against comparison.
  5. Follow the heavy buyers. Group customers by how often they bought the product before the change and track each group's repeat purchase. A loss among the most frequent buyers is the Diet Pepsi pattern; a loud protest with steady repeat buying is the Irn-Bru one.
  6. Keep a way back. Coca-Cola and PepsiCo both reversed by bringing the old version back alongside the new. Decide in advance what result would trigger that, and how long to wait for it; how long to wait before judging a change helps set the date.

Run this way, each reformulation adds to a customer world model, so the next recipe decision starts from how your own customers took the last one. For reading the reviews around a single change, see how to tell whether a product release changed your reviews.

What would Dimension Labs tell a brand team before a reformulation ships?

Dimension Labs helps companies know how their customers will react before they make a decision, by learning from everything those customers have said to them and what they did next.

For a brand or menu team, each part of the platform covers a step above. It is fully omni-channel, so a customer who says the product tastes different counts the same in a care call, a chat, an email, a retailer review or a survey. It integrates customer channels with operational data, joining those conversations to orders, loyalty or retailer sales by market, so the pilot markets can be measured against markets that kept the old recipe, with a range; that join and effects with confidence intervals are on Prove ($2,999 a month) and up. It enriches every interaction with Dimensions written for your product, such as "noticed a taste change", "objects to the announcement" or "switched to another product", that product, marketing and consumer care can all use; custom Dimensions start on Discover ($999 a month). And its visual analytics layer turns the result into a deck, a dashboard, a PDF or executive report, or an interactive data story for the leadership team that signs off the change.

Consumer care teams already use it this way. Concentrix runs chatbot programs for a large retail client across orders, returns and loyalty, with every interaction enriched for intent, journey stage, escalation triggers and resolution; in the words of Jenny Burr of its Speech Science and Analytics team, "The partnership with Dimension Labs has been instrumental in supporting our improvements in self-service performance and customer journey optimization." (Concentrix story)

When something simpler is the better choice: a single restaurant changing one dish can ask its regulars and watch the dish's sales for a month. Dimension Labs pays off when the product sells across many stores or retailers, when customers talk about it in several channels, and when you need to know whether the people complaining are the people who buy.

FAQ

Why did New Coke fail?

Its taste tests, with nearly 200,000 consumers, favored the new formula but did not measure customers' attachment to the original; by June 1985 hotline calls had risen from 400 to 1,500 a day, and the original returned as Coca-Cola classic after 79 days.

Do customers notice when a recipe changes?

Often not, if the change is small or gradual. Kraft sold more than 50 million boxes of its new macaroni and cheese recipe before announcing it, and a randomized trial found a 25% sodium cut in bread, made in weekly steps, went unnoticed.

Should you tell customers about a recipe change?

Tell them after they have tried it, and be ready to answer the ones who ask. Research shows that describing a food as healthier can lower how much people expect to enjoy it, which is why Kraft waited until the new recipe had sold for months.

What is the best way to measure customer reaction to a reformulation?

For a brand selling across many stores or markets, Dimension Labs: change the recipe in a few markets first, label every customer contact about the product, and on Prove join those customers to repeat purchases to measure the change against markets that kept the old recipe, with a range. A single restaurant can start with the dish's sales and its regulars' comments.

Export your care contacts or reviews with sales for the markets where your last product change went out, and start with one file to see how your customers took it. Every plan from Explore to Scale starts with a 14-day free trial.

Dimension Labs/Start with one file From $299 a month

Stop reporting. Start understanding.

Start with one file. Ask why a number moved. Plans from $299 a month, no demo and no contract on monthly plans.