TomTom's Third Act: Cutting 200 Jobs to Sell Maps to AI Agents

A shrinking top line, a fresh three-division split, and an MCP server bet that location data becomes the substrate of the agentic web.

TOM2.AS · Earnings Call · 2026-10-15

The third reorganization in fifteen months

On its surface this was an unremarkable quarter: group “revenue was EUR 129 million, a 6% decrease versus last year” — Taco Titulaer, CFO · 2026-10-15. Automotive operational revenue fell 19%, consumer revenue dropped 24% on the continued PND decline and memory-supply constraints, and free cash flow excluding restructuring was a thin EUR 8 million. Net cash of EUR 239 million with no borrowings remains the balance-sheet comfort blanket. Gross margin ticked to 90%, and the operating result was a EUR 6 million profit. The real content, though, was structural. CEO Mike Schoofs “reorganized TomTom around three divisions” — Mike Schoofs, CEO · 2026-10-15 — automated driving, location intelligence for enterprises, and navigation — while cutting roughly 200 roles. This is the second reorganization in roughly fifteen months, after the ~300-role cut under prior CEO Harold Goddijn in mid-2025, and it carries a ~EUR 15 million Q4 charge against ~EUR 21 million of annualized savings. Management narrowed full-year revenue to EUR 505–530 million and lifted the ex-restructuring operating margin outlook to around 7%. The framing is that navigation "remains the backbone" while growth comes from the other two units. That is a subtle admission: the segment generating most revenue is not the segment being sold as the future. Selling a reorg twice in a year is a tell — either the first cut didn't go far enough, or the growth that was supposed to arrive didn't.

Agents, MCP servers, and a Microsoft that moved house

The freshest thread is TomTom's attempt to sell location data into the Agent economy. Management launched a Maps MCP server in the Claude connector directory and expanded its Microsoft work into Azure, Fabric and Foundry. Schoofs' explanation is worth quoting at length because it is the whole thesis in one sentence:

What we do now is to build those capabilities with an MCP server that is a channel, a pathway to tap into your data that gets accessed by an agent.
Mike Schoofs, CEO · 2026-10-15

This is genuinely new for the company. In TomTom's own keyword trajectory, "MCP server," "location intelligence for enterprises," "Agent," "ecosystem," and "monetize" all appear as top gainers this quarter. Globally, agentic themes are converging across other reporters — agentic AI at HCLTech, AI Force, physical AI at Wipro — so TomTom is riding a broad wave rather than inventing one, but its angle (geospatial data as an agent-consumable product) is a rare, differentiated position for a small-cap. Note also that the Microsoft relationship was quietly reframed: the old collaboration was on the Bing side; the new one is "a new territory" on Fabric and Foundry. And yet — the honesty is striking — monetization is explicitly not here. “it does not happen overnight” — Mike Schoofs, CEO · 2026-10-15, Schoofs said, describing a sequence of credibility, then traction, then revenue. "Monetize" ranks as a gainer precisely because investors are being asked to underwrite that sequence with nothing to bank yet.

The enterprise that won't grow

Why does the AI pitch matter so much? Because the enterprise business has disappointed for years. On the Q3 2025 call, as analysts pressed on flatlining enterprise growth and a large customer building its own map, then-CEO Harold Goddijn conceded “it's a bit slower. You're right, and it's also a bit slower than I had hoped” — Harold Goddijn, Chief Executive Officer (CEO) · 2025-10-14. That customer contract phase-out is still dragging today; CFO Taco Titulaer says it will take "another three quarters to be fully phased out." So the "location intelligence for enterprises" banner is, in effect, a rebrand bolted onto an old wound — the same segment, a new label, with AI offered as the catalyst that navigation and government deals never quite became. The automated driving division houses the bigger prize: the Volkswagen/CARIAD win. But the same recurring tension persists — 2027 growth is projected to come mostly from navigation SOPs, while order intake tilts toward automated driving, whose revenue is two to three years out. On the mapless debate, Schoofs repeats the argument Goddijn made in April: “it makes self-driving technology easier when you do have a map and more reliable and redundant” — Harold Goddijn, CEO · 2026-04-16. Management is consistent here, which lends it credibility.

A caution from the tape

TomTom is a ~EUR 600 million market-cap software name with no price-tape history served in this context, so the trade backdrop is thematic. And the theme cuts both ways: over the trailing 90 days the market has been aggressively de-rating AI infrastructure — data centers, co-packaged optics, and high-bandwidth memory clusters all down 5–7%. TomTom is a data-and-software attach rather than a capex entity, so the linkage is sentiment, not mechanics — but the timing is awkward: the company is pivoting its narrative toward AI just as the market starts punishing AI exposure. The quarter's quiet logic is that TomTom is betting its future on selling dynamic data and maps as raw material for other people's agents, while its core revenue shrinks. If the agentic-access thesis lands, this is the cheap option on a genuinely rare asset. If it doesn't, it is the second reorganization that still didn't fix the enterprise line. The market will get its first real signal on "traction" in the next two quarters — and management has conspicuously refrained from telling us what that signal will look like.