You don't need ten AI subscriptions. You need the right tool for the task in front of you. Here's the short list I actually reach for, matched to the thing that's eating your time.
Takes 5 minutes to read, two weeks to prove. You need: one job that annoys you weekly. You end with: one tool you actually use, instead of five you pay for.
| The time-sink | Reach for | Why this one |
|---|---|---|
| Turning meeting notes into action items | A transcript tool + ChatGPT | It pulls out who does what and when, so nothing lives in your head |
| Writing emails that sound like you | Claude | It matches your tone best when you paste an example |
| Quick answers you can trust | Perplexity | It shows its sources instead of making you guess |
| Reading a long document fast | Any chat tool | Paste it in and ask for "the 5 things that matter" |
| Remembering to follow up | An automation, not your brain | Set it once. Let it nudge you instead of you tracking it |
| Rough first drafts of anything | ChatGPT or Claude | Getting to a bad first draft is 80% of the battle |
Start on the free tier, always
Every tool in that table has a free version that's enough to find out whether it fits your week. Two weeks of real use tells you more than any review, and you'll know exactly which limit you hit — messages, uploads, speed — which is the only useful reason to upgrade.
Pay when a specific limit is genuinely getting in your way. Not before, and not because a paid plan feels more serious.
The power move: work out which tool you're missing
Rather than reading comparisons, let your own week decide. Paste in the last two weeks of your task list, or the output of the Time Audit guide, and ask:
That last sentence is the important one. Nine times out of ten the answer is a tool you already have and haven't set up — and no comparison article will ever tell you that, because nobody writes articles recommending you buy nothing.
The other direction: what to cancel
Once a quarter, go the other way. List every subscription you pay for and ask:
Tool collecting is expensive twice — the money, and the low-level guilt of the ones you're not using. Cancelling one thing you don't use is exactly as valuable as finding one that works.
How to know it's working
Two weeks in, ask yourself one question: do you reach for it without deciding to? If you still have to remember it exists, it hasn't earned a place, no matter how good it looked in a demo. Drop it and try the next row.
Real adoption feels like the tool disappearing into the task. You don't think "I'll use AI for this," you just do the thing and it's faster.
When it goes sideways
"I signed up for four and use none." The usual pattern, and the fix isn't discipline — it's picking one row and deleting the other apps from your phone for a fortnight.
"It's good but I keep forgetting to use it." Attach it to something you already do: the tool opens when the weekly meeting ends, or when you sit down Sunday night. New habits stick to old ones, not to intentions.
"The free tier keeps cutting me off mid-task." That's the signal to pay, and it's a good one. You now know precisely what you're buying.
"My team won't adopt it." You went too wide. Pick the one person with the most painful recurring task, get it working for them, and let them tell the others. Team adoption is one convincing person, not one announcement.
"I picked a tool and the recommendations have all changed since." They will keep changing. The row you picked matters more than the brand in it.
Make it yours
- Client service: start with the email row. It's the highest-volume writing you do and the one where sounding like yourself matters most.
- Studio, shop, scheduled business: start with the follow-up row, automated. Missed follow-ups cost more than slow emails.
- Solo: start with rough first drafts. When there's nobody to bounce things off, the blank page is the real bottleneck.
- Small team: start with meeting notes to action items. It's the one that gives time back to everybody at once.
Do this next
Pick the row that matches your worst week, and use that tool for the next two weeks on real work. Put a note in your calendar for two weeks out that says: am I reaching for this without thinking?
How to use this: Don't sign up for all of these. Pick the one row that matches your biggest weekly headache, use that tool for two weeks until it's a habit, then add the next. Most people collect tools and use none. One tool, one job, done well.
(Tools shift over time, the point isn't the exact brand, it's matching one tool to one recurring task instead of forcing everything through a single app.)
Next guide: The Time Audit if you're not sure which row is really costing you the most — one prompt ranks them for you.