Money
When $5,000 is enough
Tobiloba Odejinmi · 18 Jun 2026 · 6 min · 911 words
Direct answer
$5,000 is enough when you have one written process, a person who will own the exceptions, and tools I can plug into without a new stack. You get a live workflow, a write-up, and a handover. You do not get thirty days of support, a sequenced roadmap, or several processes. If you need those, The week at $7,500 or several workflows at $15,000+ is the honest buy.
- $5,000 buys one process done, not a seat and not a program.
- It is enough when the steps are already clear and the miss is survivable.
- It is not enough when you need monitoring cover or a second workflow.
- A good $5,000 project still has token, reviewer, and integration costs after handover.
When is $5,000 the right number?
When you can point at a pile and describe the steps without a workshop. When the work is repetitive and a miss is visible. When someone on your team will read the exceptions without being chased. That is a starter. I like starters. They tell the truth fast.
I started on Fiverr. Small, clear jobs. The same instinct applies. If the job is not clear, a bigger invoice will not make it clear. It will only make the fog more expensive.
What does the one workflow have to look like?
A trigger you can name. A result you can check. Tools that already hold the data. A person for the cases that do not fit. Support answers you already give. A CV pile you already screen. A follow-up you already write. Fields you already copy between two systems.
If the workflow is 'make the company more AI', it is not a $5,000 job. It is a sentence that needs to be broken. I will help you break it on a call. I will not take $5,000 to automate a slogan.
- One owner after handover
- One primary tool path, not five new logins
- A miss you can explain without a legal team
- Volume that is real enough to learn from in a month
What will $5,000 not cover?
It will not cover me watching production for a month. It will not cover a knowledge store you have not scoped. It will not cover a second process that shares a customer record and a lot of edge cases. Those are why The week is $7,500 and several workflows start at $15,000.
It will also not cover tokens, reviewer hours, or a connector that breaks in October. Those are yours in every tier. The starter just makes that split more obvious because the build line is smaller.
When should you skip $5,000 and wait?
Wait if nobody owns the process. Wait if the data is a mess and the first week would be janitorial. Wait if a miss on this path would be a patient, a payment, or a bound policy you cannot unwind. In those shops I would rather map first, or start on a safer neighbor process.
Wait if you are buying to catch up with a competitor’s press release. The 9% Frontier Firm club is not entered through a starter invoice. Buying $5,000 of capability so you can say you did AI is how you waste a week of everyone’s time.
What should look true thirty days later?
The workflow still runs. The owner can explain a miss. Inference is a number you have seen. Reviewer time is not a surprise. A few hours have left the original pile, or you can say honestly that they have not, and why.
That is a successful $5,000. It is not a case study with a fake payback multiple. If we later move you to The week or a second workflow, we do it because the first one earned the next conversation.
How is this different from The week?
The week is the complete seven-day job. Same first process, more spine. Monitoring. Thirty days of support. A plan for what is next. Proper APIs. I tell people to book it when the process matters enough that week three cannot be a shrug.
$5,000 is enough when you want the smallest honest ship. I respect that. Just do not ask the smallest honest ship to carry a company transformation. That is a different boat, and most companies are not in it yet.
Questions people ask
When is $5,000 the right price?
When the process fits on a page, the tools exist, and one owner will live with the output. Internal field-copying, a first-pass screen, a follow-up note. Work your team already knows how to do by hand.
What does $5,000 not include?
A thirty-day support window, a plan for the next automations, a knowledge store, and the monitoring package that sits in The week. It also does not include a second workflow.
Is $5,000 cheaper than The week in practice?
On the invoice, yes. If you then pay me or someone else to add monitoring and a next-step plan a month later, maybe not. Buy The week when you know you want that spine.
Can $5,000 produce ROI?
It can produce capability ROI immediately if we ship. It can produce realized ROI if hours actually leave the desk and quality holds. It will not produce strategic ROI. Do not ask a starter build to do that.
What if the process is messier than we thought?
We stop and say so. I will not burn the $5,000 pretending one workflow is three. We either shrink the process or we talk about a different tier.
Written by
Tobiloba Odejinmi
Head of Engineering at 10mg Health. I have run engineering at Zeeh Africa and sold Insurpass and Shopl. I still write the code. If you have one process that still runs on people copying things, we can look at it in thirty minutes.
