The Catalog Built for Compliance, Not for People: Fixing the Enterprise App Store Nobody Uses

Somewhere in your organization sits an internal app catalog. It was launched with a town hall announcement, a slide deck, and a mandate: employees should be able to discover and request software on their own, without opening a ticket. And somewhere else — in the same organization — an analyst in accounting is emailing the IT service desk for the fourth time this month, asking if someone can 'get her set up with that reporting tool the finance team uses.'

The catalog, in other words, is a ghost town. Industry surveys consistently place internal marketplace adoption in the low single digits, and anyone who has audited one of these portals knows the pattern: a handful of listings last updated during the previous administration, search that returns nothing useful, and a request workflow that funnels the user right back into the ticket queue the marketplace was supposed to eliminate.

The failure is not mysterious. It is a design problem, and it is fixable.

You Built an Inventory. People Wanted a Store.

Most internal app catalogs were assembled by IT asset management teams, and it shows. The organizing logic is the CMDB: every application documented, categorized, and assigned an owner. That is exactly the right structure for an audit. It is exactly the wrong structure for a shopping experience.

Consider what a consumer app store — the App Store, Google Play — gets right that most enterprise catalogs get wrong. Search works, because metadata is maintained. Listings are current, because publishing is easy and stale apps get pulled. Reviews and star ratings exist, so a user can tell at a glance whether the tool their colleague raved about is actually any good. And critically, checkout takes seconds: tap, authenticate, done.

Now map that against the typical enterprise portal. Search depends on naming conventions nobody follows. Listings describe applications the way a procurement officer would ('License tier: FTE-concurrent, seat-managed'). Reviews don't exist, so the only social proof lives in hallway conversations the catalog can't capture. And the 'get it' button opens a request form with eleven required fields, routing to an approval chain that may take two weeks — at which point the employee has already emailed someone they know and gotten the software installed by Friday.

Every one of those friction points is a decision someone made, usually in the name of governance. Governance is legitimate. But when governance is baked into the front door rather than the back end, it doesn't reduce risk; it just reroutes traffic around the front door.

The Shadow Workaround Is Your Product Roadmap

Here is the uncomfortable reframing: when employees bypass your marketplace, they are not being noncompliant. They are telling you what the marketplace should have been.

The emails to IT are demand signals with full context attached — who wants what, for which workflow, with what urgency. The informal 'can you hook me up with a license' exchanges between colleagues are the review system your catalog lacks, just unrecorded. If your self-service portal is failing, the fastest diagnostic is to trace where the actual requests go. Map the shadow path, then ask which specific step in the sanctioned path caused the detour. It is almost always one of three things: they couldn't find it, they couldn't tell if it was right for them, or they couldn't get it quickly once they did.

Each of those maps to a concrete fix.

A Playbook for Self-Service That Survives Contact with Employees

Rebuild search around how people describe problems, not how vendors name products. Nobody searches for 'enterprise resource planning module, financials.' They search for 'expense report tool' or 'the thing the Denver office uses for time tracking.' Invest in synonyms, curated tags, and — increasingly — semantic search powered by language models, which has quietly become one of the more practical AI applications inside the enterprise.

Make listings answer the buyer's actual questions. What does this do? Who else here uses it? What does it cost my department? How long until I have it? A listing that answers those four questions in plain English outperforms a page of procurement metadata every time.

Let users rate and review. This is where governance teams flinch, and where they should stand down. Internal reviews of sanctioned software carry no security risk — the apps are already approved. What reviews do provide is the discovery signal that drives adoption everywhere else in software. A three-star average with honest comments is more useful than a polished listing nobody trusts.

Move approval behind the click, not in front of it. Pre-negotiate the request path for standard roles and license tiers so that for the majority of cases, the marketplace really is one click. Route exceptions — elevated access, unusual license volumes — to human review, but stop making every employee pay the exception tax.

Assign an owner with a retail mindset. Catalogs die when nobody is accountable for freshness. Treat the marketplace like a storefront: prune dead listings quarterly, surface new arrivals, and measure the metric that matters — the share of software requests that complete through self-service without a ticket.

The Leadership Angle

For executives, the dead internal marketplace is more than a UX embarrassment. It is a leading indicator of how digital transformation actually lands inside the company. Organizations that ship consumer-grade internal experiences — and this one is achievable within a quarter — send a signal that ripples well beyond IT: here, the tools we build for ourselves meet the standard we set for customers.

Organizations that don't send a different signal. Employees learn that 'self-service' means a form that disappears into a queue, and they behave accordingly.

The app store you built isn't empty because employees prefer email. It's empty because it was never really a store. Rebuild it as one, and the traffic follows.