Ottendo

The Ottendo method

From a named problem to a measured result.

Every initiative runs through five phases: Discover, Diagnose, Design, Deliver and Evolve. Each one ends with a short sign-off gate, so you always know what was decided and nothing moves on until you agree.

At a glance

One method for every initiative.

Scroll through the five phases: the hook, the goal and the gate you sign off before we move on.

  1. Phase 01

    Discover

    Name the problem before anyone names a tool.

    Understand the problem this initiative targets, who it affects and what success looks like, well enough to decide it is worth pursuing.

    Sign-off gate

    • Problem and success measure are agreed
    • The process owner is named
    • You agree it is worth diagnosing
  2. Phase 02

    Diagnose

    Measure how the work really runs today.

    Measure how the process works today and find the real causes of the problem, so the design fixes the right thing.

    Sign-off gate

    • Baseline is recorded
    • Root causes are agreed
    • No unresolved data or compliance blocker
  3. Phase 03

    Design

    Simplify first. Then decide what AI does.

    Decide exactly what will change, which tool will do it, and how success will be measured, before anything is built.

    Sign-off gate

    • Design and target are approved
    • Any new tool cost is approved
    • Who builds what, by when, is agreed
  4. Phase 04

    Deliver

    Prove it on real work, then hand it over.

    Build the change, prove it works on real work, and get the people who use it comfortable running it.

    Sign-off gate

    • It is used on real work and meeting its target
    • Users are trained
    • You can run it without Ottendo
  5. Phase 05

    Evolve

    Keep it working. Feed the next win.

    Confirm the initiative keeps delivering, improve it from real use, and feed what was learned into the next initiative.

    Sign-off gate

    • Stable, expand or retire: every review ends with a decision

Phase by phase

What happens, what you give, what you get.

  1. 01

    Discover

    Name the problem before anyone names a tool.

    Understand the problem this initiative targets, who it affects and what success looks like, well enough to decide it is worth pursuing.

    What we do together

    1. 1Restate the initiative as a problem, in your words.
    2. 2Identify who owns the process and who does the work today.
    3. 3Agree what success looks like: time saved, errors cut, faster response, more sales.
    4. 4Check fit: is AI the right tool, or would simplifying the process be enough?
    5. 5Note constraints: budget, tools you already own, data sensitivity, deadlines.

    What you give

    • The process owner's time
    • Examples of the problem in action
    • Any limits on budget, tools or data

    What you get

    • An initiative brief: the problem, the owner, the success measure, the constraints

    Sign-off gate

    • Problem and success measure are agreed
    • The process owner is named
    • You agree it is worth diagnosing
  2. 02

    Diagnose

    Measure how the work really runs today.

    Measure how the process works today and find the real causes of the problem, so the design fixes the right thing.

    What we do together

    1. 1Walk through the process with the people who do it: steps, handoffs, tools, volumes.
    2. 2Record the baseline: time per task, cycle time, errors or rework, cost.
    3. 3Review real examples to find patterns in what goes wrong.
    4. 4Check the data a solution would need: where it lives, how clean it is, who can grant access.
    5. 5Flag risks: customer data, privacy, compliance, what happens if the AI gets it wrong.

    What you give

    • Time with the people who do the work
    • Sample data or real examples
    • Access to the tools involved

    What you get

    • A current-state process map
    • Baseline metrics
    • Root causes, risks and blockers

    Sign-off gate

    • Baseline is recorded
    • Root causes are agreed
    • No unresolved data or compliance blocker
  3. 03

    Design

    Simplify first. Then decide what AI does.

    Decide exactly what will change, which tool will do it, and how success will be measured, before anything is built.

    What we do together

    1. 1Remove steps, double entry and unneeded approvals before adding AI.
    2. 2Design the new process: what the AI does, what people still do, where a person reviews.
    3. 3Choose the tool: what you already pay for first, then off-the-shelf, custom builds last.
    4. 4Prototype the riskiest part on sample data with 2 to 3 users.
    5. 5Set the target against the baseline and agree how it will be tracked.

    What you give

    • The process owner and future users to review the design
    • Decisions on tools and any new spend

    What you get

    • An initiative design: today's steps, the new steps, the tool, the human check, the target

    Sign-off gate

    • Design and target are approved
    • Any new tool cost is approved
    • Who builds what, by when, is agreed
  4. 04

    Deliver

    Prove it on real work, then hand it over.

    Build the change, prove it works on real work, and get the people who use it comfortable running it.

    What we do together

    1. 1Build the solution as designed, in short steps with a demo at each.
    2. 2Test on real examples, including edge cases and what happens when the AI is wrong.
    3. 3Pilot with one or a few users before rolling out widely.
    4. 4Train your team and leave a short how-to or recorded walkthrough.
    5. 5Hand over: who owns it, how to change it, who to call when it breaks.

    What you give

    • Users with time to try it and give honest feedback
    • Logins to the tools involved
    • A quick decision-maker during the pilot

    What you get

    • The working solution, in daily use
    • Pilot results against the target
    • How-to notes and a handover

    Sign-off gate

    • It is used on real work and meeting its target
    • Users are trained
    • You can run it without Ottendo
  5. 05

    Evolve

    Keep it working. Feed the next win.

    Confirm the initiative keeps delivering, improve it from real use, and feed what was learned into the next initiative.

    What we do together

    1. 1Track the result against the baseline and log the saving.
    2. 2Watch quality and cost: wrong outputs, complaints, tool spend.
    3. 3Tune prompts, rules and settings from feedback.
    4. 4Swap in better tools or models when they clearly pay off.
    5. 5Capture lessons and add new ideas to your backlog.

    What you give

    • Usage data or a simple record of time saved
    • Feedback from the people using it

    What you get

    • A wins log entry showing the result
    • A change log of improvements
    • New initiatives on the backlog

    Sign-off gate

    • Stable, expand or retire: every review ends with a decision

A worked example

An email-review skill for a business owner

An owner spends an hour a day checking staff emails to customers before they go out. One initiative fixes that.

Baseline
60 min a day
Target
< 15 min a day
  1. 01

    Discover

    The owner names the problem: about an hour a day reviewing outgoing emails. Success means under 15 minutes a day with no drop in quality.

  2. 02

    Diagnose

    Ten recent emails and the owner's corrections show most fixes are tone, missing prices and missing next steps. Baseline: 60 minutes a day.

  3. 03

    Design

    The owner's rules become a review skill for the AI assistant the business already uses: tone guide, required details, a checklist and examples. The owner still approves anything sent to a new customer.

  4. 04

    Deliver

    The skill is tested on real drafts, then the two staff who write the most emails are trained and the rules are tuned from their first results.

  5. 05

    Evolve

    Review time is tracked for a month and the saving logged. New rules are added as gaps appear, then the next initiative is picked, such as quote follow-ups.

The method stays the same whether you are a two-person shop or a team of two hundred. What changes is how fast each phase runs.

See how it scales across plans

Fix-It Call

Ready for Discover?

45 minutes with an Ottendo consultant to find where AI can save you the most time and money. It is free, with no obligation.