The Signs a Martech Tool Is About to Be Sunset (and What to Do About It)

Researching this year's round of tool comparisons turned up the same story twice, in two completely unrelated product categories. Adobe stopped selling Adobe XD as a standalone product and confirmed it's in maintenance mode - no new features, existing users only. Separately, Drift, once the standard-bearer for B2B conversational marketing, was put on a public sunset path by its owner, with customers referred to a different company's product entirely. Neither of these happened overnight, and neither was a surprise to anyone who'd been watching the signals. That's the real lesson worth pulling out of both stories: a tool's decline is visible well before the sunset announcement, if you know what to look for.

Why This Matters More for Martech Than Most Software Categories

A marketing stack is unusually exposed to vendor risk compared to, say, an internal ops tool nobody outside the company ever sees. A chatbot, a design tool, an email platform - these are often wired into a live customer-facing workflow, sometimes with years of historical data (segments, flows, brand assets) sitting inside them. When one of those tools gets wound down, the cost isn't just a subscription cancellation - it's a migration project, a data export scramble, and often a gap in whatever function that tool was performing while the replacement gets stood up. The earlier that risk gets spotted, the more it looks like a planned transition instead of a fire drill.

The Signals That Actually Precede a Sunset

An acquisition that doesn't obviously fit the acquirer's roadmap

Not every acquisition is bad news for a product - some genuinely get more investment afterward. But when a company is acquired by a competitor or a much larger platform that already has an overlapping product, or when the acquiring company's own messaging talks about the acquired team's technology or talent rather than the product itself, that's a meaningfully different signal than an acquisition framed around growing the acquired product's own customer base. Drift's path ran through two ownership changes - acquired by Salesloft, which then merged with Clari - before the sunset was announced; each transition was a checkpoint where the product's future got a little less certain, well before the final decision was made public.

A slowing release cadence

Products that are actively invested in tend to ship visible, dated changelog entries at a fairly steady pace. A product that goes quiet - months between updates, changelog entries that are entirely bug fixes with nothing new - is often already in the position Adobe eventually confirmed publicly for XD: maintained, not developed. This is checkable directly, without needing insider information - most vendors publish a changelog or release notes page, and a pattern of silence there is a visible, public signal.

Sales and marketing language that shifts from the product to something else

When a company's own public messaging starts talking more about its "platform" or its "AI agent" than about the specific product a customer originally bought, that's often an early tell that internal investment priorities have already moved elsewhere, even before any formal announcement. Intercom's rebrand around its AI agent, ahead of its own pending acquisition, is a case of a company visibly repositioning around what it believes its future actually is - worth watching in either direction, since it can also signal genuine reinvestment rather than decline, but it's a signal worth reading closely either way.

No published pricing, or pricing that only goes one direction

A product that has quietly moved from published self-serve pricing to "contact sales for a quote" isn't automatically dying, but it's worth noting as a change in how the vendor wants to manage the relationship - often correlated with a product being repositioned toward fewer, larger accounts rather than a broad customer base, which is sometimes a precursor to deprioritising the smaller end of that base entirely.

A Concrete Example

Take a mid-sized marketing team that had built a chatbot-driven lead qualification flow on a B2B conversational platform, wired into their CRM and feeding directly into their sales routing logic. The tool worked well for a couple of years. Then its parent company was acquired by a larger sales engagement platform, and the acquiring company's own public materials started talking almost exclusively about its combined revenue-intelligence roadmap, with the original chatbot product mentioned only in passing. Release notes, which had shipped something new most months, went quiet for an extended stretch. None of this was hidden - it was sitting in public changelog pages and press releases the whole time.

The team that had been watching those signals started evaluating alternatives quietly, months before any sunset was ever announced - exporting their conversation flow logic, documenting their routing rules in a vendor-neutral format, and shortlisting two replacement platforms. When the sunset announcement eventually came, with a named successor platform pointed to by the outgoing vendor, the team already had their own independent shortlist and a data export in hand. The migration still took real effort, but it was a planned project with a timeline they controlled, not a reactive scramble against someone else's deadline.

What to Actually Do About It

Don't wait for the announcement to plan an exit path

By the time a sunset is officially announced, the vendor has usually already made the decision internally for a while - the announcement is a formality, not the start of the process. Treating the early signals above as a prompt to at least sketch out "if this tool disappeared, what would we do" is a cheap exercise that pays for itself the one time it turns out to matter.

Keep your own data exportable, always

Regardless of any sunset risk, a tool that makes it hard to export your own data - segments, historical records, creative assets - is a risk on its own merits. Confirming export capability before a crisis, rather than during one, is a basic piece of stack hygiene that also happens to be the single most useful thing to have already done if a real sunset does hit.

Weight new commitments by vendor stability, not just features

When evaluating a new tool for the stack, a feature comparison alone misses a real input: how stable is this vendor's ownership and investment trajectory right now. A tool that's slightly behind on features but backed by a company clearly still investing in it is often a safer multi-year bet than a feature leader that's mid-acquisition or already showing the release-cadence and pricing signals described above. This isn't a reason to avoid every tool that's ever been acquired - most acquisitions don't end in a sunset - but it's a factor worth actively weighing rather than ignoring in favour of a pure feature checklist.

Build a light annual review into the stack, not just at renewal time

A quick annual pass across every tool in the stack - checking changelog activity, ownership status, and any public news since the last review - catches drift long before a renewal deadline forces the question. This is a small recurring habit, not a major project, and it's exactly the kind of check that would have surfaced both of this year's situations well ahead of the public announcements.

The Broader Point

Neither Adobe XD's freeze nor Drift's sunset was a secret held tightly by the companies involved - both were visible in advance to anyone tracking release activity, ownership changes, and public positioning. The actual failure mode isn't that these signals are hard to read; it's that most teams only look at a vendor's health at the moment a competitor's sales rep brings it up, or when a renewal notice lands. Treating vendor stability as a metric worth checking on its own schedule, independent of any specific crisis, is what turns an eventual sunset announcement from a scramble into something that was already mostly planned for.