What to look for · Chapter 04
AI consulting vs implementation vs managed
Agencies package their work in different ways. Some only advise. Some only build. Some run it for you long-term. Knowing the types helps you buy the right thing. It also stops you paying for what you don't need.
“Efficiency is doing things right; effectiveness is doing the right things.”
AI consulting vs implementation vs managed
There are three main engagement types. Each solves a different problem. Read across to find your fit.
| Type | What it is | Best when | Cost pattern |
|---|---|---|---|
| Consulting | Strategy and roadmap | You need direction first | Day rate or fixed fee |
| Implementation | Building and integrating | The plan is clear | Fixed project |
| Managed | Running and support | You want it run for you | Monthly retainer |
The types differ in how much the agency does. Consulting leans on your team. Managed leans on theirs.
Many buyers use all three over time. Enter wherever your gap is.
AI consulting: strategy and roadmap
Consulting answers one question: what should we automate, and how? You get a plan, not a working system. It suits teams unsure where to start.
What you get
- An opportunity map of what to automate
- A prioritised roadmap
- Rough ROI estimates
- Tool and platform recommendations
Typical length is one to four weeks. Cost is a day rate or a fixed discovery fee. Choose it when the opportunity is unclear. Skip it when you already know the workflow and outcome.
AI implementation: building the solution
Implementation turns a plan into a working automation. You get a built, integrated and tested system. It suits teams with a clear brief.
What you get
- A working, integrated automation
- Connections to your existing tools
- Documentation and a handover
- Testing before go-live
Typical length is two to six weeks per workflow. Cost is usually a fixed project fee. Choose it when the scope is clear. Skip it when you're still unsure what to build.
Managed AI automation: running it for you
A managed service keeps the automation working after launch. You get monitoring, fixes and improvements over time. It suits teams without in-house maintainers.
What you get
- Monitoring and alerting
- Bug fixes and updates
- An agreed service level (SLA)
- A regular performance report
It runs on a monthly retainer, often $3,000–$20,000. Choose it when the automation is business-critical. Skip it when you can maintain a one-off yourself.
The hybrid path: strategy plus build
Most real engagements blend the three. A common path is a short discovery, then a build. A retainer can follow for support. A fractional AI lead can guide strategy part-time. This mixes outside expertise with your own control.
How to pick the right engagement
Start from your biggest gap. Missing a plan? Start with consulting. Have a plan but no builders? Choose implementation. Built it, but nobody maintains it? A managed service fits.
For small teams
Skip standalone consulting if the scope is clear. A fixed-scope build gets you a quick win.
For enterprises
A discovery or consulting phase de-risks a large rollout. Then move to implementation and a managed service.
Common mistakes when choosing an engagement
Key takeaways
- Consulting gives direction; implementation gives a system.
- Managed services run and improve it after launch.
- Pick the type that fills your biggest gap.
- Blend them — discover, build, then retain for support.
Looking for a specific service?
Browse agencies by service — consulting, implementation or managed.
Browse by serviceFrequently asked questions
What's the difference between AI consulting and AI implementation?+
AI consulting is strategy — what to automate and how. AI implementation is the building — developing, integrating and deploying it. Consulting gives you a plan; implementation gives you a working system. Many agencies do both, but some only advise or only build.
What is a managed AI automation service?+
A managed service runs and improves your automation after launch. The agency monitors it, fixes issues and updates it over time. You pay a monthly retainer instead of a one-off fee. It suits teams that lack the staff to maintain automation in-house.
Which AI automation engagement type should I choose?+
Choose consulting when you need direction first. Choose implementation when the plan is clear and you need it built. Choose managed when you want someone to run it long-term. If you need both direction and delivery, pick an agency that does both.
How long does an AI automation engagement take?+
Consulting usually takes one to four weeks. Implementation runs two to six weeks per workflow. Managed services are ongoing, on a monthly retainer. A full journey — plan, build, then run — can span a few months.
Do I need AI consulting or can I go straight to a build?+
If you already know the workflow and outcome, you can go straight to implementation. Consulting helps when the opportunity is unclear or the roadmap is missing. A short discovery phase often bridges the two. Skip standalone consulting if the scope is obvious.
Is a retainer or a fixed project better for AI automation?+
A fixed project suits a one-off build with clear scope. A retainer suits ongoing work, changes and support. If you'll keep adding automations, a retainer is usually better value. For a single, well-defined build, a fixed project is simpler.
Can one agency handle strategy, build and support?+
Yes — many full-service agencies do all three. That gives you one accountable partner from plan to running system. Check they have real depth in each, not just breadth. Some firms specialise, so match their strength to your biggest gap.