Anyone at headquarters who is responsible for alarger network of independent locations, whether franchise partners, dealers,branches, sales agents or members of a cooperative group, knows the picturethat the year-end review often shows. The existing marketing portal is wellstocked with templates, and there are usually matching co-op advertising fundsavailable on demand. Yet only a small share of locations uses these offeringsregularly and successfully, while a large share uses them only occasionally or,worse still, not at all.
In partner marketing, this pattern has beenshowing up for years, usually as a Pareto distribution in which roughly onefifth of partners generate most of the activity. Projects in which a great dealof work has gone into rollout and onboarding show that this distribution can bechanged. At Mitsubishi Motors Germany, where we introduced our new LocalyzerCoop system in February, more than 70 percent of dealer partners were activeafter go-live. This was made possible by pilot partners who were involved earlyon, a dealer day, training sessions, webinars, tutorials and ongoing supportthat closely accompanied the rollout for months.
This effort was certainly justified and, aboveall, it worked. At the same time, it can only be transferred to organizationswith hundreds or thousands of locations to a limited extent over the long term,and it has to be repeated with every new topic, every system extension andevery new partner. That is why one question keeps us busy, and we now considerit one of the most important in local marketing: Who actually takes the firststep, the location or the system?
Where the impact gets lost
What headquarters provides is used too rarely onthe ground. As a result, visibility is missing exactly where purchasingdecisions are made, which ultimately means fewer inquiries, fewer customers andless revenue. Headquarters has invested in infrastructure and still oftenmisses its targets, because real impact only comes with actual use.
The will is certainly there among the locations.In a March 2026 survey by Uberall and Dialog among 352 US franchise brands andfranchisees, 93 percent of franchisees described local marketing as critical totheir success. Yet in a 2025 study by Constant Contact and Ascend2, franchiseesnamed lack of time (34 percent) and lack of marketing expertise (31 percent) astheir biggest hurdles alongside limited budgets.
This matches our own experience and ourconversations with individual locations. Whoever runs a kitchen showroom, aninsurance agency or a franchise business does marketing on the side. Withoutroutine, every additional step becomes a hurdle, so any task without immediateurgency gets postponed until it eventually disappears in day-to-day business.
The obvious reaction of many headquarters isthen to provide even more: more templates, another tool or a new module. Thesoftware industry knows all too well how unproductive this path is: theanalytics provider Pendo evaluated usage data from 615 software products andfound that around 80 percent of features are rarely or never used.
Why AI alone does not yet solve the problem
Generative AI now seems to offer a simplesolution, since texts, images and complete ads can be created in seconds. Forthe individual location, however, little changes at first as long as it has tocall up the AI itself, feed it with information and, above all, evaluate theresult. An AI feature that the partner has to operate actively is, for them,initially just one more task.
On top of that, AI is still far from a givenamong small and mid-sized businesses. According to KfW Research, 20 percent ofGerman SMEs use AI, with smaller businesses lagging well behind, and many localpartners belong to exactly this group.
So the decisive question remains: Who takes thefirst step? The Pendo analysis also shows that prompts from the system changeusage behavior: products that used at least 25 short in-app hints and guidesincreased the number of features used daily by 25 percent. For local marketing,this means that a system that passively waits for the location to find timewill, unfortunately, often wait in vain at most locations.
How a local marketing system should work today
In our view, local marketing systems need toevolve from a collection of tools into a kind of operating system for localmarketing, a “Local Marketing OS” that knows each location, thinks ahead and,above all, reaches out on its own. The foundation is an individual plan foreach location that connects central themes and budget rules with local factorssuch as purchasing power, competition, seasonality and past marketingperformance, is updated regularly and follows the guardrails set byheadquarters.
This gives the partner on site two new ways intothe system. The first is a dialogue in which they can ask questions the waythey would ask an experienced marketing consultant: “Which campaign should Irun now?” “What is my competition advertising right now?” “Which topics willposition me best for the year-end business?” The answers are based on their owndata, their market and the guidelines set by headquarters.
The second goes further, because it also reachesthe locations that would never ask questions on their own: the system gets intouch with concrete suggestions for current occasions. For example, with a notethat a competitor in the region is currently advertising a particular topicheavily and that a different offer would be the right response now. Or with amessage that new social media posts are available that match the location’sfocus areas exactly. Or with the observation that the campaigns on Meta havelost effectiveness, together with two new ad suggestions that can be usedimmediately.
In every case, the decision remains with thepartner, who receives a well-founded suggestion, adjusts it if needed or simplyconfirms it. Where a local campaign today can easily take 10 to 60 minutes,from opening the portal through template, budget and channels to approval, aready-made suggestion reduces the effort to a few minutes. In some cases, thepartner does not even need to log in and can confirm the suggestion directlyfrom a message. And for a location without a marketing routine, exactly this simpleaccess decides whether it takes action at all.
What changes for headquarters
Open standards such as the Model ContextProtocol (MCP) now make it possible to connect marketing systems directly to AIassistants that many employees already use. Anyone who previously had to filterdashboards and analyze exports can now ask their system and organizationquestions directly: “Which partners are most successful, and what do they dodifferently from the rest?” “Based on all campaigns from the past months, whichtemplates should we create for the year-end business?” “How are our brand andour competitors positioned, and what should our partners promote as a result?”This extends all the way to concrete tasks such as creating a new template fromexisting materials.
These questions immediately reveal the actualeffect. As long as an analysis takes days, it is done once a quarter, and forquick analyses people use whatever the dashboard happens to offer. Once aquestion, and above all an answer including interpretation, takes only seconds,decision-makers ask about what they actually want to know. Here too, the systemcan go one step further, run analyses on its own and present the results toheadquarters with context, for example as a weekly performance report on campaignsor as an alert as soon as usage declines.
For headquarters and partners to trust such asystem, however, we consider two principles indispensable. The AI only seeswhat the respective user is allowed to see in the system with theirpermissions. And every change it prepares, such as a budget shift with figuresand reasoning, is approved by a person before it takes effect. Especially inGermany and across Europe, where expectations for data protection andtransparency are rightly high, these principles determine acceptance.
A sober look
The global research and advisory firm Gartnerwarns, not without reason, that many of the AI agents currently on offerdeliver little measurable value, and estimates that only a small share ofvendors are actually developing agentic solutions. Good onboarding and thepersonal relationship between headquarters and location remain prerequisitesfor any activation. What AI can do is ensure that the support that today mostlyreaches the already active locations in person reaches every location. Thisworks best when AI features are deeply and intuitively embedded in the systemand familiar structures remain in place.
For managing decentralized networks, this puts ametric at the center that is still missing from many reports: the share oflocations that are actually active and running campaigns locally.
There is no finished blueprint for this yet,even though the first building blocks are already in use at Localyzer. What isclear to us, however, is that local marketing systems must be measured bywhether they take the first step, at every location and especially at thosethat have not had time for it so far.
Sources
- Uberall and Dialog: Closing the Franchise Gap: Local Marketing in the Age of AI, survey of 352 US franchise brands and franchisees, March 2026
- Constant Contact and Ascend2: franchise marketing study, 503 respondents in the US and Canada, January 2025, reported by MediaPost
- Pendo: The 2019 Feature Adoption Report
- KfW Research: Use of Artificial Intelligence in German SMEs, Fokus Volkswirtschaft No. 533, February 2026 (in German)
- Gartner on “agent washing”, reported by Staffing Industry Analysts and PR Wire
- Pareto pattern in the channel: Localyzer white paper “Die aktivierten 20 %” (in German)
