2 tables · 10 columns
Workflow process map template
One row per handoff: who does it, what goes in and out, the system, touch time and wait time.
This automation opportunity assessment template scores each candidate workflow on rule clarity, data, system access, exceptions and cost of a mistake, then sets the score against hours spent. It is for teams with a long list of ideas and budget for one or two. Score together, compute the weighted score and shortlist the top pick.
The register is the sheet you fill in. The scoring guide and decision rules below it are reference. Run the top pick through the AI readiness assessment next.
Automation opportunity scoring template
Free to copy. No email and no download.
CSV and Sheets copies hold the fill-in tables. The Markdown copy includes the guides.
One row per workflow. Columns E and K are formulas, shown in the steps below. Score every criterion from 1 to 5, where 5 is the easiest to automate.
| WorkflowVerb and object. | OwnerWho runs it today. | Items per monthFrom a system report. | Minutes per itemFrom a timed sample. | Hours per monthItems x minutes / 60. | Rule clarity1 to 5. Weight 25%. | Data and inputs1 to 5. Weight 20%. | System access1 to 5. Weight 15%. | Exceptions1 to 5. Weight 20%. | Cost of a mistake1 to 5. Weight 20%. | Weighted scoreScore x weight, summed. | DecisionFrom the decision rules. |
|---|---|---|---|---|---|---|---|---|---|---|---|
Scroll sideways to see every column.
Read the anchors aloud before scoring. Use 2 and 4 when a workflow sits between two anchors.
| Criterion | Weight | 1 means | 3 means | 5 means |
|---|---|---|---|---|
| Rule clarity | 25% | People decide case by case, and no two do it alike. | A written procedure exists, but people bend it. | The rules fit on a page and everyone applies them the same way. |
| Data and inputs | 20% | Inputs are paper, images or free text, and outcomes are not recorded. | A mix of structured fields and documents. Outcomes are partly recorded. | Consistent fields or documents, and past cases have the correct outcome recorded. |
| System access | 15% | No API or export. Screen access only. | Exports, or a vendor API after an IT request. | API access and a test environment are ready. |
| Exceptions | 20% | More than half the items need special handling. | Between a quarter and a half of items hit an exception. | Fewer than one item in ten needs special handling. |
| Cost of a mistake | 20% | A wrong result is costly or hard to reverse, such as a payment or a legal commitment. | Mistakes are caught at the next step within days. | Mistakes are cheap, and a person checks before anything leaves the building. |
Scroll sideways to see every column.
Starting points. Change the thresholds if you like, but set them before you score so the result is not tuned afterward.
| Weighted score | Hours per month | Decision |
|---|---|---|
| 4.0 or higher | In the top half of your list | Pilot first. Measure it with a timed sample before you commit. |
| 4.0 or higher | In the bottom half | Quick win if the build is small. Otherwise park it. |
| 3.0 to under 4.0 | In the top half | Fix the lowest-scoring criterion, then rescore. |
| 3.0 to under 4.0 | In the bottom half | Park until volume grows. |
| Below 3.0 | Any | Not yet. Standardize the process or start recording outcomes first. |
Scroll sideways to see every column.
6 steps, in the order that saves rework.
Ask each team for the tasks they would hand over first. Write them as verb and object, such as Reconcile bank statements. A dozen is a good list, and there is no need to filter yet.
Take items per month from a system report and minutes per item from a timed sample of real work. If a number is an estimate, say so in the Decision column and move on.
Do it with the process owner in the room, using the anchors in the scoring guide. Score every workflow on rule clarity before moving to data. Scoring a whole row at once lets a good first impression lift every number.
With the register header in row 1, hours per month in E2 is =C2*D2/60. The weighted score in K2 is =F2*0.25+G2*0.2+H2*0.15+I2*0.2+J2*0.2. Copy both down.
Sort by weighted score, then read each row against the decision rules. Pick one or two workflows. A score is a screening tool, so it ranks candidates and does not prove a business case.
Run the readiness assessment on it, then time a real sample and test on past cases. That is what a workflow audit does, and the sample audit shows what the output looks like.
A made-up case, so you can see what a filled-in version looks like. None of the numbers are real.
Illustrative. Volumes, minutes and scores are made up. Weighted scores use the weights in the scoring guide.
| Workflow | Hours per month | Scores (rule, data, system, exceptions, mistake) | Weighted score | Decision |
|---|---|---|---|---|
| Process supplier invoices (1,800 a month at 6 min) | 180 | 5, 4, 4, 4, 4 | 4.25 | Pilot first |
| Reconcile bank statements (600 a month at 10 min) | 100 | 4, 4, 3, 4, 3 | 3.65 | Fix system access and the cost of a mistake, then rescore |
| Book meeting rooms (300 a month at 3 min) | 15 | 5, 4, 4, 4, 5 | 4.45 | Quick win if the build is small |
| Redline sales contracts (80 a month at 45 min) | 60 | 2, 3, 3, 2, 2 | 2.35 | Not yet. Write a clause playbook first |
Scroll sideways to see every column.
How to read it. Invoices rank first because they combine the highest hours with clear rules. Room booking scores higher on feasibility but saves 15 hours a month, so it only wins if the build is cheap. Contract redlining has real hours but no consistent rules to automate, so the first job there is a playbook, not a system.
A structured way to compare candidate workflows before you build anything. Each workflow gets a feasibility score from a few criteria and a size from the hours it takes, and the two together decide which one goes first.
All the ones people name, usually around a dozen. Scoring takes a few minutes each once you have volume and minutes per item. A short list is easy to rank, and a long one shows you which ideas never needed a meeting.
Yes. They are a starting point. If a mistake is very costly in your business, raise the weight on cost of a mistake. Decide the weights before you score, so nobody adjusts them to rescue a favorite.
No score settles it. A score of 4.0 or more with high hours makes a good pilot candidate, and it still needs a measured sample and a test on past cases. We usually treat the score as a way to choose what to measure, not as the business case.
The readiness assessment takes one workflow through twelve questions across six areas and gives a band and three fixes. This template compares many workflows side by side so you can choose which one to assess.
2 tables · 10 columns
One row per handoff: who does it, what goes in and out, the system, touch time and wait time.
3 tables · 8 columns
Questions, gold answers, sources and expected behavior, plus a run log for every release.
1 table · 9 columns
A master list of every close task with owner, reviewer, due day, evidence and automation fit.
40 questions · 7 columns
Forty neutral questions to ask any AI vendor, on data, models, approvals, logging, testing and exit.
1 table · 35 items
What to settle before signing: outcome, workflow, data, test thresholds, security, delivery and support.
A 30-minute call with an engineer. You will leave knowing what a system could do for that workflow and what it would take to build.
What happens next
Strategy call
We learn how the workflow runs today and what a good outcome would be.
Workflow audit
We map the process and measure where a system would save the most time.
Proposal
A written scope, timeline and fixed price for your first system. No obligation to build.
Scenox Studios AI assistant