What I'd Look At Before Renewing Anything

The vendor renewal calendar at most big companies follows a familiar rhythm. Did the volume arrive, did the dashboard match the contract, did performance hold up, is the budget there. Sign here.

Of course, the actual evaluation has always gone further. Has the risk profile changed? Is anyone getting too comfortable? Have other options emerged with better ROI? Can we build or insource part of what we're paying for?

What's changing now is how the buy/build/switch equation gets answered. Timing varies by product and industry. But some of those conversations are already starting to change.

A Sharper Line

The reason is structural. The platform layer that vendors have sold for a decade - the software, the workflow, the user-facing system itself - is becoming something a competent team can build in a week with today's models and APIs. This isn't going to happen overnight, but it's going to happen. Some companies and clients will move faster than others, for plenty of reasons. Every vendor is going to face two questions: does what they sell still justify what we pay, and does how they sell it still work for us?

AI is going to sharpen the line between vendors with real value and vendors without it. The distinction existed before. It's about to become a dividing line - between vendors who survive the AI revolution and vendors who don't.

One camp built businesses where the user-facing wrapper was the whole differentiation. A clever interface around generally available capability. Light data. No proprietary loop. Those companies are in trouble. The bodies stay upright for a while because of brand, habit, and multi-year contracts, but their days are numbered.

The other camp built businesses with something durable underneath. Data and learning accumulated over years. Business logic refined from operating in a specialized space over time. Capabilities that came from working across many customers - the kind of cross-cutting view no single customer can build alone. Niche expertise that took real investment to develop. Those vendors are fine. The format they've been selling will change, which will be uncomfortable, but the substance is intact. In some cases more valuable than before, because the noise around it is clearing.

Persado, Specifically

One company I think is squarely in the second camp - and well poised to make the pivot - is Persado. They've been a Verizon partner for years and delivered good value the whole way. My team carried their budget for part of that run. The work in plain English: brief them on a campaign, they generate copy variations, they score those variations against their model of what tends to convert for your audience, you get measurable conversion lift. They're particularly strong in regulated verticals - financial services, telecom, retail - where compliance and conversion both have to land in the same draft.

Underneath the product is something harder to replicate. More than a decade of campaigns with real outcomes attached - over 100,000 of them, by their own count - in industries where the feedback is unambiguous. Their own homepage puts it this way: "Any LLM can generate a first draft - that's 10% of the problem." Argue with the exact percentages, but the spirit is right. The other 90%, in their telling, is performance prediction, compliance, and learning from real outcomes.

The product Persado has sold for a decade is a platform. Customers log in, brief the system, get copy back. That layer is the most exposed to LLM commoditization. As customers get more used to LLMs embedded in their workflows - or running the workflow entirely - patience for manually operating someone else's interface runs out. They'll point their own agents at the platform and have them crawl, scrape, and operate the site. Inefficient, fragile, slow - for now. But that's how customers will want to operate.

The Format Has to Change

For vendors in the second camp, the work is to find new ways to deliver what they have. What they actually know - the data, the patterns they've learned, the business logic refined over years - needs to reach customers as something other than "log into our software." An API customers can call from inside their own systems. Custom-trained models that draw on the vendor's data without exposing it. Their scoring built directly into customer tools. Secure shared environments where vendor data and customer data can be combined without either side handing over what's underneath.

The pricing model has to shift too. Per-seat licensing made sense for a platform people log into. The new offerings don't fit that. They need pricing tied to actual use - per query, per training session, per project. Different way to charge, different forecasting on both sides, different line on the customer's budget.

None of that is a small change. Different sale, partly different buyer, different deployment, different conversation with the customer, different invoice. And it all has to happen while the existing platform business is still funding the company. The technology is the easy part.

The underlying asset is defensible. Nobody shortcuts their way to years of accumulated data, learning, and refined logic in a specialized space. Customers don't take it with them when they switch vendors. Newcomers can't reverse-engineer it from public sources. The moat compounds with every new customer, every new use, every additional year of operation.

It's a fundamentally different position from thin-value AI vendors. They face an existential question. Vendors with real substance face a packaging question.

From the Buyer Side

Which brings me back to the renewal calendar.

Vendor evaluation is starting to look different. Some contracts you'll renegotiate, because the vendor is still doing real work but the format they're selling it through is mispriced or no longer what you need. Some you'll walk away from, because the substance was always thin and you can replicate the wrapper internally for a fraction of the cost. Some you'll renew with quiet enthusiasm, because the vendor is sitting on something AI can't replicate, and you'll start asking those vendors for that something delivered differently than before.

The vendors who come through this period well are the ones whose value was real before AI, with leadership clear-eyed enough to repackage that value for what's coming. Come renewal time, look past what a vendor says about AI. Look at the work showing up underneath: real refocusing, real repackaging, real preparation for a customer base about to operate completely differently.