AI Guides › Trend Watch

The Difference Between A Feature And A Product

By Nigel Guy · 3 min read

A new tool launches doing one clever thing, and it's natural to judge it only on how well it does that thing. What that skips is a more useful question: is this a product — something that can stand on its own and keep going — or a feature that a bigger tool you already use could add in a single update? Build a workflow on a feature-shaped tool and you may find it absorbed, discontinued or made redundant sooner than you'd planned for.

The rule: before relying on a new tool, ask whether it would still make sense if the platform underneath it, or the big tool next to it, shipped the same thing — if not, treat it as a feature and don't build anything load-bearing on it.

The Feature-or-Product Test

Question Leans product Leans feature
What does it add beyond the underlying model or platform? A substantial workflow, data or integration layer Mostly a well-phrased instruction and a nice interface
Does it hold anything you'd miss? Your history, shared workspace, team setup, stored work Nothing persists between uses
Could a major tool add this in one update? Hard — it needs specialised depth Easily — it would be a button
How does it make money? Clear pricing that matches the value Unclear, or dependent on someone else's pricing staying put
Who is it for? Specific people doing a specific job Everyone, for everything

Three or more answers in the right-hand column and it's a feature, whatever the website calls it.

Why it matters to you

Feature-shaped tools aren't bad. Many are genuinely useful and cheap to try. The risk sits in dependency: templates, automations, colleagues trained on it, work stored inside it. The more of those you build up, the more exposed you are if the feature gets folded into something else or simply stops.

Using a feature-shaped tool well

  1. Use it for work with no long tail. One-off tasks, drafts, conversions — things where you take the output and leave.
  2. Keep your inputs and outputs elsewhere. Don't let it become the only place your work lives.
  3. Note how it does what it does. If it's essentially a clever instruction pattern, write the pattern down so you could reproduce it in another tool.
  4. Revisit it at your next tool review. Check whether the big tools have caught up — see "The Honest Case For A Quarterly Tool Audit, Not A Weekly One".

What to skip

Skip paying annually for anything that fails the test. Skip the "first mover" argument from the tool's own marketing — being first to a feature doesn't protect it. And skip dismissing a tool just because it's built on someone else's model; nearly everything is. The question is what's added on top.

Guardrails

All 751 AI guides · JulieMango plans from £17/mo