How Mars Stopped Losing Good Products After Launch

For years, Mars watched strong products lose momentum after launch.

The ideas were good and the prototypes impressed stakeholders, but by the time a product reached the region meant to sell it, the people had changed and the context had been lost.

Cristina Garcia-Cervigon, SVP Growth Strategy, Portfolio and Innovation, spent two years rebuilding how Mars innovates to fix that. The gum is the first proof it works.

Five years as a GM at Mars gave Cristina a clear view of where the company was struggling. The front end was strong: teams could spot a consumer need, build it into a real proposition, and get it ready for market. The problem came after launch, in the phase Cristina calls the nurturing gap.

Once a product was on the shelf, it needed people, money, and fast iteration to keep improving until it won. Mars kept falling short there. Products would launch with momentum, then drift as the organization moved on to the next thing.

This is exactly where small companies beat big ones. A challenger launches with a handful of SKUs, throws everything into the launch, and keeps iterating on what sells for months. In a big company, the people move to the next thing, the money thins out, and the product stops improving before it finds its market. The challenger stays on the same few products long after the incumbent has moved on, and eventually wins the sales.

Cristina traced this gap to three specific causes inside Mars - all coming from how the work is organized:

All three causes she found to be structural, which meant they could be redesigned. And that is what Cristina spent the next two years doing.

Cristina's first move was to fix what the company aims at. Understanding consumers well is the starting point, but a team that understands consumers and then chases every opportunity doesn’t go very far. So Mars connected that consumer understanding to a smaller question: out of everything consumers want, which areas will the company actually go after?

Mars answered it with a study of what it calls demand spaces: a map of the snacking landscape that shows how consumers behave and where their unmet needs sit, with trends laid over the top. The map gave the company a shared read of the territory. Leadership then used it to set priorities, “the big rocks worth chasing”, so that the portfolio backs a defined set of consumer areas rather than every idea that surfaces.

This is what Cristina calls strategic clarity. When leadership has already agreed on which consumer areas matter, each team knows which pain points to own, which projects deserve to be built, and the regions know what they are signing up to.

With the demand spaces set, the structure underneath them is where Cristina focused on next.

As part of the redesign, Mars split innovation into two teams with different jobs. The first is a small Discover team, with around 40 people: R&D prototypers, insights, and marketing, working in weekly agile cycles. Its job is to own consumer pain points and build solutions for them without caring which brand or category the answer lands in. This ensures the solutions are based upon fair judgment, instead of a preference towards a particular product.

Discover produces more than just finished products: prototypes, a packaging direction, a shift in product architecture are part of this team’s responsibility. Importantly, a large part of what this team does is educating themselves about a consumer area that will not become a Discover product at all. The findings flow straight to the regions and improve what they already sell, or they route to one of Mars's other divisions where it fits better than it would in a new launch. One team studying a pain point ends up feeding the whole company - which is the opposite of a category team building only what its own brand can sell.

Once Discover has a real project worth building, the work moves into a pod; the Develop stage. A pod is led by an intrapreneur, which in Mars means a senior operator who runs the project like a small general manager: they have enough experience to break through the obstacles that stall innovation inside a big company, and enough authority to keep every function moving at pace. The pod carries the project from pre-launch direction to in-market growth. The pod unit is larger than Discover, because it pulls in supply, design studios, and the regions.

The two-team structure removes the handover. The same people stay involved as a project moves from Discover into Deeclop, and the regions sit in from the first pain point, with frequent check-ins across the whole arc. There is no single moment of transfer for ownership to fall through. The pod is built for the phase big companies usually drop: the months after launch, when a product needs people, money and fast iteration to win.

AI runs on both sides of the launch. At the front end, concepts, iterations, packaging, and communication that used to take weeks can be generated in a day, so Discover can test many directions before a pod commits to one. After launch, the same teams use AI to track how a product is performing in market and how the project itself is moving, letting a pod understand whether a launch is working - and act on it while the window is still open. Cristina is clear that the tools only matter when they sit inside how people already work: a tool no-one opens changes nothing.

Before the redesign, Mars measured every project the same way from day one, applying financial targets even in the weeks when the goal was still to learn whether the idea worked.

Cristina’s new approach takes the pressure off teams to prove revenue and encourages them to pay attention to the learnings they made in those crucial early experiment stages: she splits the KPIs by stage. Early on, in Discover and Develop, teams are measured on what they learn. Financial measures arrive later, once a product is meant to deliver. Her phrase for this discipline is to stop things "at the right moment and to do it fast.”

Underneath the stage-based KPIs is a shift from single projects to platforms. A team owns a platform, which is a consumer area the company has decided to win in, with several projects running inside it. Any one project can fail without the platform failing, so no individual has their name tied to the survival of one bet.

This change does not just “switch on” with a new scorecard. Cristina is frank that the work is mostly cultural - and it's slow. When a project stops, the people who worked on it would still rather skip the meeting where it happens. That’s why her work focuses not on “the what” but on “the how” a team experiences a change.

Ever since these changes, speed is improving: short agile cycles put the right people in the room often enough to accelerate what is working, reprioritize, and stop what is not, at a pace Cristina says the company could not manage before. And every project stopped early returns money and attention to the ones earning it.

Under the old model, the global center would build a product and hand it to a region to sell. Cristina flipped the sequence. The regions come in on day one and help shape the product as it develops, so by launch they already own what they helped build.

The commitment is set up front. Leadership aligns first on the consumer pain points and the priorities worth chasing, then on exactly which projects enter the pods. Cristina describes this early agreement as almost contractual: the regions are signed up to the projects before the building starts, rather than asked to adopt them after.

She is realistic about why this has to happen. Anything long term in a large organization competes with the demands of today, and the region carrying today's number will rationally choose today’s product.

So now, her answer is to protect the work: ring-fenced funding kept separate from the core, so the money for new products stays put even when the core business needs more.

Committing resources on day one also means going outside for the parts Mars is slow at. Cristina explains it with a formula, momentum equals mass times velocity. Mars already has the mass: its size in snacking, its people in the regions, its brand and category knowledge. What it lacks is velocity; plain speed. The same setup that makes Mars good at running its big brands makes it slow at the early, experimental work a new product needs.

So Cristina looked at the process and asked a simple question: where is Mars strong, and where does it slow itself down? Then, she brought in a partner for each slow step. For early R&D and testing, Mars works with Cargill and Mission Kitchen. For getting to market, it works with Walmart and Pilot Lite. The pods still own the innovation: the partners just take the slow steps off Mars's hands so the whole thing moves faster.

Cristina is open that this takes some swallowing of pride. To bring in outside help, a company has to admit where it is slow, which is hard for an organization that prides itself on running the core. But Mars does it anyway, since they are aware of the benefits outside partnerships bring.

Retailers are one of those outside partners, and the way Mars organizes its work makes them easier to bring in. Mars does not pitch a retailer a single product. It commits to winning a whole consumer area, then runs several products inside it, and brings the retailer in on the area. Backing an area a company is serious about is a much easier yes for a customer than backing one product on its own. This is how Mars built its mood-management gum, a line aimed at stress relief and staying focused, with Tesco. Rather than taking a finished product to Tesco to stock, Mars brought the retailer in at the very start, as a partner in the whole area, and the two worked on it together from the start.

One consumer area that came out of the demand spaces study was mood management: the everyday need to reset, calm down or stay focused. Leadership made it a priority.

Discover started from two facts. People already chew gum to change how they feel, and the target consumer cared about natural ingredients. Fast testing turned the area into a pipeline of several products rather than one bet. The gum was one of them. It moved into a pod, the launch region was committed from the start, and Tesco worked on it with Mars from the beginning.

Cristina names three reasons it worked: a real understanding of the consumer, genuine speed to market, and the region and customer aligned from day one.

Eight months after launch, the region that sells the gum is the region that helped shape it. On launch day it had shelf space, local budget and a team pushing it, because the people who built the product are the people selling it. It never changed hands, so it never lost momentum.

Not everything in the pipeline will land. Some projects will be stopped early; some will launch and underperform. Cristina is fine with both, because the platform keeps running when a single project inside it fails. As she puts it, "when it fails, it's not a problem. It just continues."

The gum proved a big company can get rid of the handover and get its best launch in years. The rest of the mood-management pipeline, running on the same setup, is where Mars finds out if it can do it once again.