Nobody in this category publishes a migration timeline, which is convenient for them and useless to you. Here is ours, including the parts that go wrong and the questions we ask before quoting a date.
Integrations are the timeline driver. Everything else is more predictable than you have been led to expect.
Typical phases for a catalog and search project. Durations are planning estimates that firm up after discovery, and the real variable is how many systems we connect to and how complex they are.
Weeks 1–2
Which system owns which field, what your ERP has been customised to do, what your current URLs are worth, and the real state of the product data. The audit is part of onboarding, not a pre-sales ask.
Weeks 2–4
The spec fields for each product family, the rules for converting units to one standard, and which system wins when two disagree. This is the work that decides whether search behaves.
Weeks 3–8
Always the longest task. Scoped first so the schedule means something. Ready-made connections shorten it; a decade of undocumented ERP customisation lengthens it.
Weeks 4–8
Runs in parallel. Your brand applied to the library of ready-made page sections, the templates built once and reused, and marketing trained to operate them before go-live rather than after.
Weeks 6–10
Existing product URLs preserved or redirected. Losing indexed product pages is one of the few genuinely expensive mistakes available here, and it is entirely avoidable.
Weeks 8–12
The new catalog running against live data before anything switches, so the failure modes appear while the old site is still there to fall back to.
After go-live
Content and campaign pages move to your team on day one. If your marketers are still filing tickets a month later, the project did not land.
Planning estimates, not commitments. Discovery is what turns them into a date we will stand behind.
Almost always. It is why the catalog audit happens in week one rather than at the point it becomes a surprise.
Some business rule encoded in 2014 by someone who left in 2017. We go looking for these deliberately in discovery.
Product URLs that encode a category that no longer exists. The redirect map is built before anything retires.
The most common non-technical failure. Someone has to be accountable for the catalog copy, and it is better decided in week two than week ten.
We know this platform well, including from the inside: a substantial amount of our work has been building search, distributor integration and BOM tooling on top of client sites running on it.
The content management story stays strong. What changes is where product data lives and what filtering it can support:
If you have invested heavily in Kentico recently, your marketing team depends on its personalisation, and the gap is specifically that buyers cannot find parts, then the right project is adding our search, distributor integration and BOM tools to your existing site rather than a replatform.
The case for a full migration is different: an ageing instance nobody wants to upgrade, a catalog layer that has become a standing engineering cost, and a marketing team that files tickets for content changes. If two of those three are true, the arithmetic usually favours moving.
A stated agenda, a real solutions engineer, and a written follow-up with what we found whether or not you proceed.