--- title: "How to study your business for AI, task by task" description: "Study real tasks before choosing AI tools. Record the work, source files, exceptions, owners, and checks so each proposed function has a clear purpose." canonical: "https://scalewithsearch.com/articles/how-to-study-business-for-ai" --- ## Site navigation - [Scale With Search](https://scalewithsearch.com/) - Services - Services - [AI systems](https://scalewithsearch.com/ai-systems) - [Site build](https://scalewithsearch.com/work#site-build) - [Whole website fix](https://scalewithsearch.com/work#whole-website-fix) - [Content library with SEO](https://scalewithsearch.com/work#content-library) - [Working session](https://scalewithsearch.com/audit) - Start here - [Send your brief](https://scalewithsearch.com/work#send-your-brief) - [Example scope](https://scalewithsearch.com/work#review-the-example-scope) - [Prices and terms](https://scalewithsearch.com/work#prices-and-terms) - Industries - Real estate - [Real estate agents and teams](https://scalewithsearch.com/for/real-estate) - [Platforms, ownership, and IDX](https://scalewithsearch.com/for/real-estate/real-estate-website-platforms-ownership-idx) - [Search, content, and AI](https://scalewithsearch.com/for/real-estate/real-estate-search-content-ai) - [Inquiries, campaigns, and handoffs](https://scalewithsearch.com/for/real-estate/real-estate-website-inquiries-campaigns-handoffs) - [Buyer, seller, and local pages](https://scalewithsearch.com/for/real-estate/real-estate-buyer-seller-local-resources) - [Rules, privacy, and trust](https://scalewithsearch.com/for/real-estate/real-estate-website-rules-privacy-trust) - Browse by business - [Trades and home services](https://scalewithsearch.com/for/#group-trades) - [Health, wellness, and animal care](https://scalewithsearch.com/for/#group-practices) - [Law, finance, and IT](https://scalewithsearch.com/for/#group-professional) - [Marketing and SEO agencies](https://scalewithsearch.com/for/#group-agencies) - [Shops and manufacturers](https://scalewithsearch.com/for/#group-manufacturing) - [Restaurants, shops, studios, and nonprofits](https://scalewithsearch.com/for/#group-local-services) - Chiropractic practices - [Chiropractic and wellness practices](https://scalewithsearch.com/for/chiropractic-and-wellness-practices) - [Chiropractic and physical therapy clinics](https://scalewithsearch.com/for/chiropractic-and-physical-therapy-clinics) - [All industries](https://scalewithsearch.com/for/) - Resources - Free tools - [AI readiness check](https://scalewithsearch.com/check) - [Free tools](https://scalewithsearch.com/tools/) - [Free Starter Kit](https://scalewithsearch.com/kit/business-memory-starter-kit.zip) - Guides - [AI-native library](https://scalewithsearch.com/ai-native) - [Industry intelligence](https://scalewithsearch.com/intelligence) - [Office job guides](https://scalewithsearch.com/guides/) - [Business memory](https://scalewithsearch.com/articles/business-memory-for-ai-agents-guide) - [Agent governance](https://scalewithsearch.com/articles/ai-agent-governance-guide-small-business) - [Migration and ownership](https://scalewithsearch.com/articles/leaving-vendor-ai-memory-guide) - Articles - [All articles](https://scalewithsearch.com/articles/) - [What changed](https://scalewithsearch.com/changes/) - About - Evidence - [Proof](https://scalewithsearch.com/proof) - Scale With Search - [About](https://scalewithsearch.com/about) - [How it works](https://scalewithsearch.com/how-it-works) # How to study your business for AI, task by task Study your business for AI by following real work from its trigger to its finished result. List each recurring task. Record who does it, what they read, where they wait, and who checks the output. Decide what should change after you have that record. Start with a completed job on someone's desk. Ask the person to reconstruct it using the actual files. A general question such as “What could AI do here?” leaves too much room for guesses. A finished job gives you something to inspect. The deliverable is a task inventory with evidence beside each row. It should describe the business as it runs today. The future design comes later. ## Choose the boundary of the study Name the part of the business you will inspect first. You might follow an inquiry through its reply, a service request through its handoff, or a reporting cycle through review. Choose a start event and an end result that people can recognize. Include the people on both sides of that boundary. If an office coordinator prepares a packet for a delivery manager, speak to both. Ask the manager what makes the packet usable. Ask the coordinator what information arrives late or needs repair. Write the purpose at the top of the inventory. For example: “Understand the work required to prepare a weekly service summary.” Avoid using “install AI” as the purpose. That describes a proposed tool before the work has been studied. The [NIST Map playbook](https://airc.nist.gov/airmf-resources/playbook/map/) recommends documenting purpose, users, context, and requirements. It also calls for considering alternatives that do not use AI. This inventory is a practical way to begin that examination. Source checked October 26, 2026. ## Collect examples before interviewing the team Ask each participant for an ordinary completed job, a difficult exception, and any written instructions they used. These are suggested evidence categories, not a required sample size. Include more cases when the work varies across customers, seasons, or service types. Keep the source files in a location the business controls. Record where each example came from and who may access it. Use a redacted copy when the study does not need personal or confidential details. Do not collect passwords in the inventory. For each example, save the input and the accepted result. Include the correction history when it explains how the job changed. A polished final document alone will not show the decisions that produced it. Label missing evidence plainly. “No example found” is a useful finding. It tells you where a proposed function would depend on assumptions. ## Ask people to show the work Use the same interview sequence for each role. Ask what starts the task, which record they open first, what they produce, and what they do when the normal path fails. Then ask them to walk through their chosen example. Pause at each handoff. Ask what the next person needs before work can continue. Find the message, field, or file that signals readiness. If someone says “I know when it is ready,” ask what they noticed in this particular case. Separate the written rule from the person's explanation. When they differ, record both and ask the responsible owner to resolve the conflict. Do not quietly turn an improvised workaround into official policy. Finish by asking which slow, low-value task they would remove first. Treat the answer as a candidate to examine. Keep their explanation beside the evidence. ## Make each row a task someone can recognize Use a verb and a concrete result for each task name. “Prepare the service summary” is clearer than “reporting.” “Check whether the request contains the required details” is clearer than “intake.” Split a row when its parts have different owners, permissions, or acceptance checks. Preparing a reply and sending it belong in separate rows. One creates an artifact. The other represents the business to someone outside it. Use a small set of columns: | Field | What to record | |---|---| | Trigger | The event that starts the task | | Responsible role | The person accountable for completion | | Inputs | Named records and their locations | | Work | The actions performed | | Output | The result handed to the next person | | Exceptions | Conditions that interrupt the normal path | | Check | How someone decides the result is usable | | Evidence | The example used to verify this row | Keep the first version readable. Add detail when it changes a decision about the work. ## Separate effort, waiting, and repair Record active work time separately from elapsed time. Ask people to label estimates as estimates. When practical, observe a run and record what happened. Do not turn a remembered duration into a measured baseline. Track waiting by cause. The task may wait for a missing document, a decision, a response, or an available reviewer. Put that dependency in its own field. It deserves a different remedy from a slow drafting step. Record repair as well. Note which output had to be corrected, why it failed, and who fixed it. Include the time spent checking and repairing when comparing possible changes. Use the inventory to ask where effort goes. Do not use it to promise a particular saving. A proposed design still needs testing against the current process. ## Find the rule behind each decision A task row needs more than a sequence of clicks. Ask what makes the person choose one branch over another. Find the price list, policy, service boundary, or prior approval that supports the choice. Record a source location and an owner for each rule. If two files disagree, mark the decision unresolved. Have the owner select the current version before the rule becomes part of an AI function. Use [the context selection guide](/articles/what-context-should-an-agent-read) to decide which records a selected function should receive. Keep the wider study evidence separate from its operating inputs. Keep facts separate from examples. An old customer packet can illustrate the process without establishing today's terms. A current policy file can establish the rule without proving that staff follow it. For the later writing stage, use [the process documentation guide](/articles/how-to-document-business-processes-for-ai). This study decides what exists and what needs attention. The guide helps turn a selected process into instructions. ## Test the inventory with a cold walkthrough Give a draft inventory row and its example to someone who did not write it. Ask them to explain the trigger, inputs, owner, finish line, and exception path. Have them point to the evidence for each answer. When they need an explanation, add the missing fact or mark the gap. Do not fill it with a plausible guess. Then ask the person who performs the work to confirm the revised row. Use the difficult case as well as the ordinary one. Check whether the inventory explains why that case took a different path. A missing exception is a study finding, even if the normal sequence is clear. Keep a short decision list beside the inventory. Give each unresolved question an owner and a next action. The study can remain useful while some tasks stay blocked. ## Show how the inventory changes a proposed build Consider a fictional service firm preparing weekly updates. Its owner asks for an AI reporting function. The study follows one update from source records to the approved message. The inventory separates collecting status records, resolving conflicts, drafting the update, reviewing it, and sending it. The manager owns conflicting status decisions. The coordinator gathers records. The owner approves the final customer message. That proposed scope is narrower and more inspectable than “automate reporting.” A first function could prepare a draft from approved records and list unresolved items. It would stop before sending. The manager would still decide which disputed status is correct. This example illustrates the study method. It is not a customer result or a claim about measured performance. ## Finish with decisions the business can use For each task, record a proposed next step: stop it, simplify it, automate a defined step, assist a person, or assign it elsewhere. Include the reason and the evidence. Keep unresolved tasks visible. Choose one candidate for a written function scope. Name the result, allowed sources, approval boundary, spend limit, alert recipient, and acceptance checks. Use [the build acceptance guide](/articles/what-a-build-acceptance-list-contains) to turn “works well” into observable checks. The [AI systems method](/ai-systems) connects the study to sorting, building, and training. Start by sending the recurring job and the source material you already have through [the work form](/work). ## Related: Ownership and Migration - [Document your business processes so an AI can follow them without a briefing](/articles/how-to-document-business-processes-for-ai) - [What Context Should an Agent Read Before It Starts?](/articles/what-context-should-an-agent-read) - [What a Build Acceptance List Contains](/articles/what-a-build-acceptance-list-contains) ---- ```text .|########||. .|########||. .|########||. |##||. .||##|. |##||. .||##|. |##||. .||##|. |#|. .|#|. |#|. .|#|. |#|. .|#|. |#| |#| |#| |#| |#| |#| .#| |#. .#| |#. .#| |#. .#. .#| .#. .#| .#. .#| |#. .#| |#. .#| |#. .#| |# ...... #| |# ...... #| |# ...... #| .# ||#########| #| .# ||#########| #| .# ||#########| #| .#. |######||######|. .#. .#. |######||######|. .#. .#. |######||######|. .#. #. .##|###|##|#|######| .# #. .##|###|##|#|######| .# #. .##|###|##|#|######| .# || |##|#||||||||||||#||#| || || |##|#||||||||||||#||#| || || |##|#||||||||||||#||#| || # |#||||||||||||||||||||#| #. # |#||||||||||||||||||||#| #. # |#||||||||||||||||||||#| #. || |#||||||||||||||||||||||#| || || |#||||||||||||||||||||||#| || || |#||||||||||||||||||||||#| || # .#||||||||||||||||||||||||#| # # .#||||||||||||||||||||||||#| # # .#||||||||||||||||||||||||#| # || ....|||#||||||##|#|||#|#||##||||....|. || || ....|||#||||||##|#|||#|#||##||||....|. || || ....|||#||||||##|#|||#|#||##||||....|. || #. . ....|# ....#||| ||| .#||| ....#. .# #. . ....|# ....#||| ||| .#||| ....#. .# #. . ....|# ....#||| ||| .#||| ....#. .# # . ||||#| .#####|| . .#| .#||| ||||# #. # . ||||#| .#####|| . .#| .#||| ||||# #. # . ||||#| .#####|| . .#| .#||| ||||# #. .| ||||| #. |#|||||. || #. |#||. .||||| || .| ||||| #. |#|||||. || #. |#||. .||||| || .| ||||| #. |#|||||. || #. |#||. .||||| || || |.... #. ....|#. |. ...|. ....|| || || |.... #. ....|#. |. ...|. ....|| || || |.... #. ....|#. |. ...|. ....|| || #. .||||||##||||||##||####|||||||#||||||#| .# #. .||||||##||||||##||####|||||||#||||||#| .# #. .||||||##||||||##||####|||||||#||||||#| .# # .||####|###########################|. # # .||####|###########################|. # # .||####|###########################|. # # .#||||||.#.|| # |. ..# #| #|||||#| # # .#||||||.#.|| # |. ..# #| #|||||#| # # .#||||||.#.|| # |. ..# #| #|||||#| # # .#|||||| .. |# ## |#| ...#|#|||#| # # .#|||||| .. |# ## |#| ...#|#|||#| # # .#|||||| .. |# ## |#| ...#|#|||#| # # .#|||#|# .# .#| #| ##|.#..#|||||#| # # .#|||#|# .# .#| #| ##|.#..#|||||#| # # .#|||#|# .# .#| #| ##|.#..#|||||#| # # .#||||||############|######|||||#| # # .#||||||############|######|||||#| # # .#||||||############|######|||||#| # # |...||#||||||#||||#|#||||||#|||||#||| |.| # # |...||#||||||#||||#|#||||||#|||||#||| |.| # # |...||#||||||#||||#|#||||||#|||||#||| |.| # #. .. ||||# .|||##|. |#| .|| || .|||#. #| # .# #. .. ||||# .|||##|. |#| .|| || .|||#. #| # .# #. .. ||||# .|||##|. |#| .|| || .|||#. #| # .# || | ...|| ...##| | #| .|. |. ##### .. .| || || | ...|| ...##| | #| .|. |. ##### .. .| || || | ...|| ...##| | #| .|. |. ##### .. .| || || ||||| || ||||#| . |. |. |#. ||||| |#| || || || ||||| || ||||#| . |. |. |#. ||||| |#| || || || ||||| || ||||#| . |. |. |#. ||||| |#| || || .# |.....#|....|#.||||.|||##.|#|....||.#||.#. #. .# |.....#|....|#.||||.|||##.|#|....||.#||.#. #. .# |.....#|....|#.||||.|||##.|#|....||.#||.#. #. #.|||||||#############################| ||| .# #.|||||||#############################| ||| .# #.|||||||#############################| ||| .# || ##||#||#||||||#||#||#||#||##. || || ##||#||#||||||#||#||#||#||##. || || ##||#||#||||||#||#||#||#||##. || # .#|||||||||#||#|||||||||||#| # # .#|||||||||#||#|||||||||||#| # # .#|||||||||#||#|||||||||||#| # || |#||||||||||||||||||||||#| || || |#||||||||||||||||||||||#| || || |#||||||||||||||||||||||#| || .# |#||||||||||||||||||||#| #. .# |#||||||||||||||||||||#| #. .# |#||||||||||||||||||||#| #. || |#||||||||||||||||||#| || || |#||||||||||||||||||#| || || |#||||||||||||||||||#| || #. .######|#|#########| .# #. .######|#|#########| .# #. .######|#|#########| .# .#. |#####||||#####| .#. .#. |#####||||#####| .#. .#. |#####||||#####| .#. |# ||########|| #. |# ||########|| #. |# ||########|| #. |# ...... #| |# ...... #| |# ...... #| |#. .#| |#. .#| |#. .#| |#. .#. |#. .#. |#. .#. .#| |#. .#| |#. .#| |#. |#| |#| |#| |#| |#| |#| .|#|. .|#|. .|#|. .|#|. .|#|. .|#|. |##||. .||##| |##||. .||##| |##||. .||##| .||########||. .||########||. .||########||. Scale With Search 2026 [scalewithsearch.com](https://scalewithsearch.com) ```