Everyone Agrees Your HR Systems Should Talk. Nobody Makes You Care.

A marketer's view of a category that solves a real problem and then describes it in language no human responds to.

I spend my time helping companies position themselves: working out what they actually do for a customer and saying it in language a human recognises rather than a spec sheet only an engineer could love. I am not going to pretend to out-engineer anyone on payroll and HR technology. There are people who have lived inside this world for thirty years and understand it far better than I ever will. That is not my job.

My job is to notice something they are often too close to see. The people who understand this technology best tend to describe it in a way that almost no buyer responds to. The technology is genuinely clever, but nobody buys clever. They buy the feeling that the thing they are responsible for is finally handled. And HR and payroll integration is, right now, one of the clearest examples I have come across of a category sitting on a genuinely valuable story and telling almost none of it.

Let me show you what I mean.

The argument everyone has, and the one that matters

There are, broadly, two camps in enterprise HR technology, and they have been arguing for years.

One camp says consolidate: put everything on a single platform, one system, one source of truth, and stop the madness of a dozen tools that do not talk to each other. The other says pick the best tool for each job, even if that means running several specialists side by side. Both are describing a real frustration. The all-in-one route can leave you with a system that is strong at the centre and merely adequate at the edges, exactly where payroll and local compliance live. The best-of-everything route can leave you with a pile of brilliant tools that each hold a different version of the truth.

Both camps are half right, which is the most annoying kind of right. The real answer is a strong core surrounded by the specialists you actually need, connected properly. The argument was never really suite versus specialist. It was always about what holds them together.

Which is the exact moment this stops being a technology question and becomes a marketing one. Because the companies doing that connecting work are sitting on the most valuable story in the category, and most of them are leaving the best of it untold. They are describing the wiring when they should be describing what it does for a human being.

What the buyer is actually worried about

Stay with the technology for one more minute, because you cannot position something you do not understand, and the detail here matters.

Even the most unified core does not do everything. The big HCM platforms are excellent at being the system of record, but deep, country-specific payroll and specialist workforce management usually sit outside them. Payroll shows it most sharply. The moment a company operates in more than one country, the tidy single-system picture falls apart, because very few providers cover every market. Some companies run several regional payroll providers side by side. Others consolidate onto one global provider, but even that provider is, underneath, stitching together local engines and partners, because no single engine handles every country's rules. The integration challenge does not vanish when you pick one vendor. It just moves inside that vendor.

Now, here is where I stop describing the machine and start doing my actual job.

Notice that none of what I just wrote is what the buyer lies awake worrying about. They are not worrying about engines and connectors and data mapping. They are worrying about whether everyone gets paid correctly and on time, in every country, every cycle. Whether a regulator finds a gap. Whether the expensive platform they championed to the board ever delivers the single, honest view of the workforce they promised it would. That is the human reality underneath all the technical language. And it is almost never the thing the category talks about.

The worry is not misplaced, either. Payroll is where this shows up most expensively, so it makes the sharpest example: running payroll in a single country is one thing, but the moment you cross borders the risk climbs steeply. By one industry estimate, around a quarter of companies running payroll in a single country have been fined for getting it wrong, and that rises to somewhere between half and two-thirds once payroll spans several countries. But payroll is only the most costly instance of a problem that runs through the whole stack. The same gap opens wherever two systems have to agree: workforce management feeding the HCM, time and attendance, benefits, several payroll engines reporting into one source of truth. Every connection you add is another place the data can quietly diverge. So the thing the buyer lies awake worrying about is not a vague anxiety. It is the most likely place for the whole arrangement to fail.

"Seamless" is the most over-promised word in the business

Ask any vendor and the connecting is always handled. It is "seamless." There is a "pre-built connector." It is "fully integrated, out of the box."

A great deal of that is integration theatre, and buyers have learned it the hard way. They have switched on the pre-built connector that covered most of what they needed and left the awkward, high-stakes remainder to manual workarounds and good intentions. They have learned that "seamless" on the slide meant "achievable, with enough budget and time" in practice. They have found that the painful edge cases, the unusual pay element, the country-specific rule, the thing that only matters once a year but matters enormously when it does, are exactly where the generic connector quietly gives up.

So the word "seamless" has stopped meaning anything. Everyone says it, nobody believes it, and the companies who can genuinely deliver it are using the same exhausted language as the companies who cannot. That is not a technology problem. That is a positioning problem. When your truest claim is also your most clichéd one, the answer is never to say it louder. It is to stop describing the feature and start proving the outcome.

You are not selling connectors. You are selling a night's sleep.

This is the heart of it, and it is the thing I would change first for any company in this space.

The people doing the genuinely hard integration work are very often the worst at selling it, and not because they are bad marketers. It is because they are experts, and experts describe the engine room. Connectors, endpoints, syncs, APIs. All true, all impressive to other engineers, and all completely beside the point for the person signing the contract.

You can hear the gap in a single line. Most companies in this space describe themselves roughly like this: "pre-built connectors to 200+ HR and payroll systems, with configurable data mapping and automated sync." Every word may be accurate. Strung together, it is word salad: no meaning, just syllables, and it lands on a buyer as noise. Now say the same capability the way a human experiences it: "your people get paid correctly in every country you operate in, and your leadership finally sees one true headcount number." Same product. Completely different value. The first sentence describes the plumbing. The second describes what the buyer truly wants, which is to stop worrying.

That is the whole move, and it is not cosmetic. Most marketing problems are positioning problems wearing a technical costume. Selling integration by listing the connectors is like selling a house by describing the foundations. Nobody is moved by concrete. They are moved by a home that does not fall down. The job is to translate, relentlessly, from what the technology is into what it means for the human relying on it.

AI raises the stakes, and it does not change the rule

AI only sharpens this. Every credible AI use case in HR depends on clean, connected data, so the value of integration goes up, not down. But the language trap is the same: "AI-ready data architecture" is engine-room talk, while "you can finally trust what your systems are telling you" is the human version. A fresh wave of impressive-sounding technical claims is coming, and the same opportunity sits underneath all of them. Whoever translates it into something a person cares about wins.

The opportunity

So here is where I land.

This is a category that has spent years arguing about architecture and describing itself in the language of the engine room, while the thing it actually delivers, continuity, compliance, and the quiet confidence that the people who keep a business running will be paid and protected, goes almost entirely unsaid. Everyone already agrees the systems should talk to each other. The opportunity belongs to whoever can prove they make that happen and then describe it like it matters to a human being, because it does.

I have spent a career convinced that there is no such thing as B2B or B2C marketing, not really. There is only ever one person on the other end of it, and everything you say is either landing with them or it is not. But knowing that is not the same as doing it, and this category shows exactly where the doing breaks down. The technology is not the problem. The translation is. Somewhere between what the platform does and what the buyer feels, the story stops being about a person and starts being about a data pipeline, and that is the precise point at which most of this category loses the plot.

Because the integrator is not the payroll engine, and it is not the HCM. It is the layer that makes them agree. Its job is to give a company one version of the truth, where a change made in one system is trusted everywhere else, where HR data and payroll data tell the same story, and where the dashboard the board is looking at is built on numbers nobody has to second-guess. The engine-room way to say that is real-time bidirectional sync across the stack. The human way to say it is this: you finally know that what your systems are telling you is true.

So the work is not to build something more impressive. It is to say, plainly, what the impressive thing already does for the person relying on it: an end to the quiet dread that two systems disagree and nobody knows which is right, and the confidence that the whole connected machine is telling you the truth. Whoever learns to say that first will not just sound different from everyone else in the category. They will be the one the buyer finally understands, and the one they finally trust.