Not sure where AI would pay off? Take the 3-minute opportunity check →
Strategy · 5 min read

Your first 90 days with a fractional tech lead

By the InTheory team · Updated November 2026

A fractional technology lead's first 90 days should follow a clear arc: diagnose the business and agree one priority in month one, deliver and test it with real users in month two, and measure adoption and results in month three. This guide covers what you should provide, expected deliverables and signs it is off track.

Key takeaways
  • Month one ends with one agreed priority, a baseline and written acceptance criteria.
  • Month two builds and tests the improvement with real examples and real users.
  • Month three measures adoption and results against the baseline and sets next priorities.
  • Your side matters: an internal owner, access, and time from the people doing the work.

Why the first 90 days matter

The first three months of any fractional arrangement set its pattern. If they produce a clear priority, a working improvement and measured results, the relationship has a basis for continuing. If they produce meetings, a long list of ideas and no change in the business, it rarely recovers.

A defined 90-day arc helps both sides. You know what to expect at each stage, and the technology lead has a reason to focus rather than spreading across every request. The pattern below is the one InTheory uses, but it is a sensible structure whoever you hire: diagnose, deliver, measure.

It also helps to agree on what the arrangement is not. A fractional lead is not a help desk for password resets or a replacement for your IT provider. Being clear on this early keeps their time on the priority you chose.

Before day one: what you should provide

Preparing these in advance can save the first two weeks. A lead who spends month one chasing logins and invoices has less time for the work you are paying for.

  • An internal owner who will meet the lead weekly and carry work between meetings.
  • A list of current software subscriptions with costs, users and renewal dates.
  • Admin or guest access to the main systems, in the lead's own named account.
  • Time with the people who do the work, typically a few hours each in the first month.
  • Two or three real examples of the workflows that hurt most, such as a recent proposal or a new client file.
  • Any constraints: regulations, client contract terms, insurance requirements or budget limits.

Days 1 to 30: diagnose

The first month is about understanding the business and choosing the right first problem. Expect the lead to interview the owner and managers about priorities, then sit with the people who do the work and watch how it happens. They should review the software stack, map one or two key workflows step by step, and identify where time is lost, data is retyped or handoffs fail.

By day 30 you should have a short written diagnosis: a ranked list of opportunities with rough value and effort, a recommendation for one pilot, a baseline for the pilot's key measures, and written acceptance criteria. For example, a home remodeling company might agree to improve estimate follow-up, with a baseline of how many estimates get any follow-up today and how long they take to close. The owner approves the scope before any building starts.

Be ready for the diagnosis to surprise you. The most valuable first project is often not the one the owner expected. A firm that came in wanting an AI chatbot may find that the bigger gain is fixing how enquiries are logged and followed up. A good lead will explain the reasoning so you can make the call.

Days 31 to 60: deliver

The second month is for building and testing the agreed improvement. Depending on the scope, that might mean configuring features in software you already own, building an automation in Zapier or Make, setting up an AI drafting step with human review, or managing an outside developer or vendor. The lead should test with real examples from your business and involve the people who will use the result.

Expect regular, short updates: what was done, what was learned, any decisions needed. Before launch, the acceptance criteria are tested and the results recorded. Training happens on real work, with a one-page guide for each role. The old way of doing the task is scheduled for retirement so staff are not running two processes.

Expect some decisions to come back to you during the build. A lead may need you to approve email wording, confirm which clients are excluded from an automation, or decide whether a step should stay manual. Fast answers keep the month on schedule, so agree how you will handle these, such as a short weekly decision list.

Days 61 to 90: measure

The third month checks whether the change worked. The lead tracks usage, such as how many staff use the new process and whether steps are being skipped, and compares the workflow measures with the baseline. Expect a second round of training and small adjustments as real use reveals rough edges.

By day 90 you should have a short results report: baseline versus current numbers, adoption levels, what worked, what did not, running costs, and a recommendation for the next priority. That report should be honest. If the pilot did not move the numbers, the report should explain why and what that teaches you.

Ask for the numbers in the same format as the baseline, so the comparison is direct. If the baseline was time per proposal measured over four weeks, the result should be time per proposal measured over a comparable period, with any differences in workload noted.

Deliverables you should expect

Some quick wins often appear along the way, such as cancelling unused subscriptions or fixing a setting that was causing duplicate data. Those are welcome, but they should not replace the main pilot.

  • A written diagnosis with a ranked list of opportunities.
  • A signed scope for the pilot, including baseline, acceptance criteria and review points.
  • A system register: each tool, its owner, cost, renewal date and admin access.
  • The working improvement, built in accounts your company owns.
  • Role-based training materials and documentation of any automations.
  • A 90-day results report with next priorities.

Signs it is off track

Watch for a few warning signs. If the first month ends without one agreed priority and a baseline, the engagement is drifting. If building starts before anyone has written acceptance criteria, expect arguments later about whether it is finished. If the people who do the work have not been consulted by day 45, adoption will suffer. If the 90-day report talks about activity rather than measured results, ask for the numbers.

Raise these early and directly. A good technology lead will welcome the feedback and adjust. At day 90, decide deliberately: continue with the next priority, change the shape of the arrangement, or stop. Any of those can be the right answer, as long as it is based on what the numbers and your team are telling you.

Common questions

How much time will my team need to give a fractional tech lead?

In the first month, expect a few hours from each person involved in the chosen workflow, plus a weekly meeting with the internal owner. Month two needs time for testing and training. Month three needs less, mainly for feedback and the results review.

What if the first pilot does not work?

That is useful information, as long as it was measured. The 90-day report should explain whether the cause was the tool, the process or adoption, and recommend whether to adjust, try a different approach or move on to another priority.

Can we work on more than one priority in the first 90 days?

It is usually better not to. One well-chosen pilot that is delivered, adopted and measured builds trust and capability. Quick wins can happen alongside it, but a second major workstream tends to dilute both.

Next step

Bring one expensive workflow. Leave with a clear next step.

In 30 minutes we look at one recurring workflow, what it costs you today and whether there is a credible case to improve it. If there is not, we will say so.

Book a 30-minute workflow review Not ready to talk? Take the 3-minute check

30 minutes with Phil or Martin. No technical brief, no pitch deck. Phil or Martin replies personally, usually within one business day.

Book a 30-minute review