Give every customer an AI account manager.
What if every customer had someone thinking about them all the time? Someone who remembered their priorities, noticed a change in their business, spotted a useful new development and arrived at the next meeting with an idea worth discussing. Give every account that kind of attention with an AI advocate working behind the scenes for your team.
Try it now
Copy the prompt into your AI tool of choice, then answer its questions as they come.
Paste it into any AI chat to start. For most people, this is all you need.
Everything on this page in one paste: the prompt plus the full plan, every step and what to check. Use it for longer moves, to plan ahead, to share with a colleague, or if your AI can't open links.
How it goes
What the AI will do
- Builds an internal AI account manager that keeps investigating what would help each customer.
- Connects customer correspondence with news about the customer, competitors and industry to prepare useful service ideas and meeting briefs.
- Tests the account memory, recurring searches and delivery of its findings to your team.
What to have ready
- A customer group or a few accounts to begin with; a rough description is enough.
- Available customer correspondence, CRM and calendar access, or permitted examples; the AI will investigate connections.
Where you come in
- Correct the account picture and decide which findings would be useful.
- Choose the monitoring cadence, internal recipients and what the AI may update automatically.
What you'll get
A tested internal account-advocacy tool with customer-specific memory, recurring research and usable briefs; unavailable connections or unattended monitoring remain explicitly unfinished.
You're done when
The agreed accounts have a usable, tested internal advocate that remembers their situation and delivers relevant research, service ideas and meeting preparation. Recurring operation and account coverage are verified or explicitly marked unfinished.
These are general instructions, not a description of this user's particular AI system, tools, access, or circumstances. Adapt the method to the user's goal and the capabilities you can actually use, while preserving the activity's essential sequence, review points, and intended result. ### Start with the activity and the user's context Read the complete activity and guide the user through it. Use the activity prompt yourself; the user does not need to paste it again. Keep the responsibilities of the user, the AI, and any colleague clear. Begin with what the user has already supplied. Establish the intended result and resolve missing context through focused questions when needed. Work within the activity's stated scope, permissions, required confirmations, and stopping points, including any deliberate order for introducing information. Make progress with the available information and reasonable, clearly identified assumptions where the activity allows them. ### Check the capabilities needed for this work Inspect relevant available tools, connectors, and data. Use small, relevant, non-destructive checks to establish whether necessary access or a capability actually works. Naming a product does not establish access to it. Let this activity determine what needs checking; do not impose a universal preliminary questionnaire or a full system audit. When useful access is missing: - Explain what you cannot reach and how that material would help. - Investigate practical setup, connection, export, or other access alternatives before asking the user to gather material or troubleshoot. - If progress depends on access, request only the specific user action or permission needed to enable it or provide the material another way. Continue useful independent work while awaiting the response. - Offer to walk the user through an export or upload. Welcome files, links, copied text, rough notes, or an interview as supported by your system and the activity, and use the best available route. If the missing access is optional, continue with what is available. ### Ask focused questions and respect the answers Ask questions only when necessary to complete the work well or satisfy the activity's required review points. Build on answers already supplied. Ask one question at a time, or a small group of closely related questions. Wait on work that depends on the answer, and continue useful work that does not. Explain unfamiliar language with examples where helpful. Confirm answers or decisions you supply on the user's behalf before treating them as the user's final answers or decisions. This does not require separate approval of every sourced finding; follow the activity's specific review requirements. ### Keep evidence and independent contributions distinct Distinguish evidence, inference, and missing information. Check your work against the source material. Use independent agents when your system actually supports delegation and independent investigation or review would improve the work: - Give each agent a bounded brief and the context appropriate to its role. - Keep discovery independent until synthesis where the activity requires it. Preserve any requirement for the user or colleagues to contribute before seeing other explanations. - Follow the activity's stated fallback and any required confirmation when independent agents or fresh contexts are unavailable. Preserve reviewer isolation where required; describe shared-context or sequential passes accurately. ### Produce, check, and revise the result Do the substantive investigation and production. Carry the activity through to the most complete, usable result within its scope, rather than ending with a plan or handing the main work back to the user. When missing information does not prevent useful work, proceed with clearly identified assumptions where the activity permits them. Check the result against the activity's completion criteria. At each substantive delivery, state what is complete and any material unfinished work, assumptions, helpful missing information, and remaining uncertainty. Explain how these affect what the user can use, and help the user revise the result. Preserve the activity's stopping point. Do not turn a rehearsal, investigation, or test-design activity into execution that the user has not authorized. Build an internal AI account manager for each customer in the agreed scope: a persistent advocate that continuously investigates what could help that customer and equips our people to act. The core work is customer understanding, proactive research and useful internal preparation. Begin with a manageable pilot, then make the design repeatable across accounts. ### Discover the customers and what matters to them Use the company, product, customer names or rough examples I supplied. Inspect and try available CRM, email, chat and other messages including WhatsApp where accessible, support history, account plans, sales and renewal records, meeting transcripts and calendar access. Search by company, named people, product, current projects and important dates. Follow references to proposals, commitments, usage evidence, service records and earlier reviews. For each account, recover its goals, business model, current use of our offering, expressed frustrations, requests, buying context, open promises and upcoming meetings. Identify the actual people, their roles, stated concerns and previous responses to our ideas. Keep direct statements, observed behavior and inferred needs distinct. Trace later changes so an old objection or cancelled project does not become a permanent account fact. The account is an organization with several people, not an invented single personality. Keep source links, event dates, last-checked dates and coverage gaps. Try supported connection or export routes before asking me to collect material. With a few supplied examples, build a provisional account profile and identify the missing context; do not claim to have investigated a complete relationship. Show representative profiles and ask for corrections that change the service we should provide. Carry those corrections with provenance into the persistent account record. ### Design the account’s ongoing research For each account, identify useful public sources and concrete queries: the customer’s own announcements, filings, product or service changes, hiring and expansion news; its relevant competitors; industry developments; and changes in our own offering that may help it. Follow news to original announcements or other directly supporting evidence when available. Record the date and distinguish an announced plan from a completed change. Connect each finding to the account’s documented circumstances. Produce a candidate insight only when you can explain what happened, why it may matter to this customer, what we could do and what is still an assumption. Explore three kinds of value: - Better service: for example, a customer expanding into a new region may need a different implementation or support arrangement. - A new way to use or think about our product or service: a changed workflow may reveal an application nobody discussed before. - Better meeting preparation: an upcoming review can include the customer’s latest development, previous commitments and a useful question. These are illustrations to adapt, not evidence about our customers. Check current account discussions before proposing an idea we already tried or a need already resolved. Avoid padding a brief with unrelated headlines. Preserve helpful absence: no consequential new development is a valid monitoring result. ### Agree the working arrangement and build it Discuss how the team actually manages accounts and what would be ideal. Use existing tools when they support the desired experience. Show a small working example of an account brief and the useful choices: a daily or weekly digest, event-driven findings, and meeting-triggered preparation. Let me choose a cadence and internal destination suited to the business. Near-continuous monitoring means a tested event or scheduled process; an ordinary chat does not keep searching after it ends. Build the account store, source connections, research jobs, evidence-linked briefs and retrieval interface. Each account record should contain the current profile, evidence, open commitments, previous suggestions and their disposition, upcoming events, last successful source refresh and unresolved gaps. Preserve updates without overwriting the source of an earlier conclusion. Use stable account identifiers and separate records so one customer’s private information does not enter another’s brief. Specify for each job its trigger, sources, account scope, useful output, delivery destination and what happens after a failed read or empty result. Verify current official documentation for the integrations actually chosen. Inspect and test what can be read, stored, scheduled and delivered before relying on it. Agree which internal updates and notifications may happen automatically and which suggestions require judgment. This activity’s ordinary delivery is to our team; any later customer outreach needs its own established authorization. If an essential connection is missing, build and run the account-research components using permitted exports or a supplied sample, prepare the exact connection request and show what remains manual. Keep the usable pilot separate from a claim of full-time operation. Help with the smallest necessary access action rather than handing the setup problem back to me. ### Test useful advocacy and ongoing operation Use representative accounts with different needs. Hold back some historical developments and messages while designing the first profiles, then replay only information available at each test cutoff. Check whether the tool links each development to the right account, produces a worthwhile service idea or meeting question, avoids repeating a handled suggestion and preserves evidence. Future knowledge must not improve a simulated earlier brief. Test a changed priority, a cancelled meeting, an outdated source, a duplicate news item, an ambiguous company name, a failed connection and an account with no useful new news. Check separation using deliberately distinguishable test-account records. Try the actual recurrence or event trigger and internal output destination where authorized; a manual run does not verify scheduling. Verify that a repeated event does not create repeated notifications or duplicate commitments. For many accounts, assign bounded account groups to investigators with complete account sources, common evidence fields and explicit completion criteria. Keep separate customer contexts where possible, track coverage and have the lead reconcile shared industry findings without mixing private records. Use successive waves or manageable saved passes when agent capacity is limited. Give an independent reviewer the generated briefs, underlying sources, customer profiles and expected behavior without your rationale. Ask whether each account would be better served by what its human manager now knows. Check unsupported needs, incorrect account matches, missed commitments and useless news summaries. Repair findings and rerun affected cases. Ask me to judge a few real briefs for usefulness and incorporate my corrections. ### Deliver the account advocates Deliver the working tool, access or launch instructions, the actual account briefs, source and account coverage, verified trigger behavior, and a simple way to correct a profile, dismiss an idea, change cadence or pause the process. Explain how an additional account is added and demonstrate that path with an authorized pilot account or test fixture. State what is running, what was tested manually and which connections still block the intended service. Finish when the agreed pilot performs useful account advocacy and can be continued through the chosen mechanism, or when a clearly identified access dependency prevents that with the functioning components delivered. Do not claim all customers have ongoing coverage until those accounts and sources are actually enrolled and checked. These instructions were written in September 2026 using the frontier models available at the time. You may now have better tools or models, or restrictions that make some tasks harder. Your AI should adapt to the tools and access you actually have and find the best available way to help. Everything here is a starting suggestion for you and your AI, not a proven best practice for every setup. Feel free to rewrite any of it to better meet your needs.
View full AI instructions
Give every customer an AI account manager.
What the AI will do
- Builds an internal AI account manager that keeps investigating what would help each customer.
- Connects customer correspondence with news about the customer, competitors and industry to prepare useful service ideas and meeting briefs.
- Tests the account memory, recurring searches and delivery of its findings to your team.
What you can provide
- A customer group or a few accounts to begin with; a rough description is enough.
- Available customer correspondence, CRM and calendar access, or permitted examples; the AI will investigate connections.
Where you'll take part
- Correct the account picture and decide which findings would be useful.
- Choose the monitoring cadence, internal recipients and what the AI may update automatically.
What you'll get
A tested internal account-advocacy tool with customer-specific memory, recurring research and usable briefs; unavailable connections or unattended monitoring remain explicitly unfinished.
Activity context
What if every customer had someone thinking about them all the time? Someone who remembered their priorities, noticed a change in their business, spotted a useful new development and arrived at the next meeting with an idea worth discussing. Give every account that kind of attention with an AI advocate working behind the scenes for your team.
For the user
Name a customer group or start with a few accounts. The AI will investigate the relationships, propose how ongoing advocacy should work, and build and test the tool with you. You supply judgment about the customer and choose where its findings should arrive.
Guidance for the AI
These are general instructions, not a description of this user's particular AI system, tools, access, or circumstances. Adapt the method to the user's goal and the capabilities you can actually use, while preserving the activity's essential sequence, review points, and intended result.
Start with the activity and the user's context
Read the complete activity and guide the user through it. Use the activity prompt yourself; the user does not need to paste it again. Keep the responsibilities of the user, the AI, and any colleague clear.
Begin with what the user has already supplied. Establish the intended result and resolve missing context through focused questions when needed. Work within the activity's stated scope, permissions, required confirmations, and stopping points, including any deliberate order for introducing information. Make progress with the available information and reasonable, clearly identified assumptions where the activity allows them.
Check the capabilities needed for this work
Inspect relevant available tools, connectors, and data. Use small, relevant, non-destructive checks to establish whether necessary access or a capability actually works. Naming a product does not establish access to it. Let this activity determine what needs checking; do not impose a universal preliminary questionnaire or a full system audit.
When useful access is missing:
- Explain what you cannot reach and how that material would help.
- Investigate practical setup, connection, export, or other access alternatives before asking the user to gather material or troubleshoot.
- If progress depends on access, request only the specific user action or permission needed to enable it or provide the material another way. Continue useful independent work while awaiting the response.
- Offer to walk the user through an export or upload. Welcome files, links, copied text, rough notes, or an interview as supported by your system and the activity, and use the best available route. If the missing access is optional, continue with what is available.
Ask focused questions and respect the answers
Ask questions only when necessary to complete the work well or satisfy the activity's required review points. Build on answers already supplied. Ask one question at a time, or a small group of closely related questions. Wait on work that depends on the answer, and continue useful work that does not. Explain unfamiliar language with examples where helpful.
Confirm answers or decisions you supply on the user's behalf before treating them as the user's final answers or decisions. This does not require separate approval of every sourced finding; follow the activity's specific review requirements.
Keep evidence and independent contributions distinct
Distinguish evidence, inference, and missing information. Check your work against the source material.
Use independent agents when your system actually supports delegation and independent investigation or review would improve the work:
- Give each agent a bounded brief and the context appropriate to its role.
- Keep discovery independent until synthesis where the activity requires it. Preserve any requirement for the user or colleagues to contribute before seeing other explanations.
- Follow the activity's stated fallback and any required confirmation when independent agents or fresh contexts are unavailable. Preserve reviewer isolation where required; describe shared-context or sequential passes accurately.
Produce, check, and revise the result
Do the substantive investigation and production. Carry the activity through to the most complete, usable result within its scope, rather than ending with a plan or handing the main work back to the user. When missing information does not prevent useful work, proceed with clearly identified assumptions where the activity permits them.
Check the result against the activity's completion criteria. At each substantive delivery, state what is complete and any material unfinished work, assumptions, helpful missing information, and remaining uncertainty. Explain how these affect what the user can use, and help the user revise the result.
Preserve the activity's stopping point. Do not turn a rehearsal, investigation, or test-design activity into execution that the user has not authorized.
Your AI prompt
Build an internal AI account manager for each customer in the agreed scope: a persistent advocate that continuously investigates what could help that customer and equips our people to act. The core work is customer understanding, proactive research and useful internal preparation. Begin with a manageable pilot, then make the design repeatable across accounts.
Discover the customers and what matters to them
Use the company, product, customer names or rough examples I supplied. Inspect and try available CRM, email, chat and other messages including WhatsApp where accessible, support history, account plans, sales and renewal records, meeting transcripts and calendar access. Search by company, named people, product, current projects and important dates. Follow references to proposals, commitments, usage evidence, service records and earlier reviews.
For each account, recover its goals, business model, current use of our offering, expressed frustrations, requests, buying context, open promises and upcoming meetings. Identify the actual people, their roles, stated concerns and previous responses to our ideas. Keep direct statements, observed behavior and inferred needs distinct. Trace later changes so an old objection or cancelled project does not become a permanent account fact. The account is an organization with several people, not an invented single personality.
Keep source links, event dates, last-checked dates and coverage gaps. Try supported connection or export routes before asking me to collect material. With a few supplied examples, build a provisional account profile and identify the missing context; do not claim to have investigated a complete relationship. Show representative profiles and ask for corrections that change the service we should provide. Carry those corrections with provenance into the persistent account record.
Design the account’s ongoing research
For each account, identify useful public sources and concrete queries: the customer’s own announcements, filings, product or service changes, hiring and expansion news; its relevant competitors; industry developments; and changes in our own offering that may help it. Follow news to original announcements or other directly supporting evidence when available. Record the date and distinguish an announced plan from a completed change.
Connect each finding to the account’s documented circumstances. Produce a candidate insight only when you can explain what happened, why it may matter to this customer, what we could do and what is still an assumption. Explore three kinds of value:
- Better service: for example, a customer expanding into a new region may need a different implementation or support arrangement.
- A new way to use or think about our product or service: a changed workflow may reveal an application nobody discussed before.
- Better meeting preparation: an upcoming review can include the customer’s latest development, previous commitments and a useful question.
These are illustrations to adapt, not evidence about our customers. Check current account discussions before proposing an idea we already tried or a need already resolved. Avoid padding a brief with unrelated headlines. Preserve helpful absence: no consequential new development is a valid monitoring result.
Agree the working arrangement and build it
Discuss how the team actually manages accounts and what would be ideal. Use existing tools when they support the desired experience. Show a small working example of an account brief and the useful choices: a daily or weekly digest, event-driven findings, and meeting-triggered preparation. Let me choose a cadence and internal destination suited to the business. Near-continuous monitoring means a tested event or scheduled process; an ordinary chat does not keep searching after it ends.
Build the account store, source connections, research jobs, evidence-linked briefs and retrieval interface. Each account record should contain the current profile, evidence, open commitments, previous suggestions and their disposition, upcoming events, last successful source refresh and unresolved gaps. Preserve updates without overwriting the source of an earlier conclusion. Use stable account identifiers and separate records so one customer’s private information does not enter another’s brief.
Specify for each job its trigger, sources, account scope, useful output, delivery destination and what happens after a failed read or empty result. Verify current official documentation for the integrations actually chosen. Inspect and test what can be read, stored, scheduled and delivered before relying on it. Agree which internal updates and notifications may happen automatically and which suggestions require judgment. This activity’s ordinary delivery is to our team; any later customer outreach needs its own established authorization.
If an essential connection is missing, build and run the account-research components using permitted exports or a supplied sample, prepare the exact connection request and show what remains manual. Keep the usable pilot separate from a claim of full-time operation. Help with the smallest necessary access action rather than handing the setup problem back to me.
Test useful advocacy and ongoing operation
Use representative accounts with different needs. Hold back some historical developments and messages while designing the first profiles, then replay only information available at each test cutoff. Check whether the tool links each development to the right account, produces a worthwhile service idea or meeting question, avoids repeating a handled suggestion and preserves evidence. Future knowledge must not improve a simulated earlier brief.
Test a changed priority, a cancelled meeting, an outdated source, a duplicate news item, an ambiguous company name, a failed connection and an account with no useful new news. Check separation using deliberately distinguishable test-account records. Try the actual recurrence or event trigger and internal output destination where authorized; a manual run does not verify scheduling. Verify that a repeated event does not create repeated notifications or duplicate commitments.
For many accounts, assign bounded account groups to investigators with complete account sources, common evidence fields and explicit completion criteria. Keep separate customer contexts where possible, track coverage and have the lead reconcile shared industry findings without mixing private records. Use successive waves or manageable saved passes when agent capacity is limited.
Give an independent reviewer the generated briefs, underlying sources, customer profiles and expected behavior without your rationale. Ask whether each account would be better served by what its human manager now knows. Check unsupported needs, incorrect account matches, missed commitments and useless news summaries. Repair findings and rerun affected cases. Ask me to judge a few real briefs for usefulness and incorporate my corrections.
Deliver the account advocates
Deliver the working tool, access or launch instructions, the actual account briefs, source and account coverage, verified trigger behavior, and a simple way to correct a profile, dismiss an idea, change cadence or pause the process. Explain how an additional account is added and demonstrate that path with an authorized pilot account or test fixture. State what is running, what was tested manually and which connections still block the intended service.
Finish when the agreed pilot performs useful account advocacy and can be continued through the chosen mechanism, or when a clearly identified access dependency prevents that with the functioning components delivered. Do not claim all customers have ongoing coverage until those accounts and sources are actually enrolled and checked.
Done when
The agreed accounts have a usable, tested internal advocate that remembers their situation and delivers relevant research, service ideas and meeting preparation. Recurring operation and account coverage are verified or explicitly marked unfinished.
These instructions were written in September 2026 using the frontier models available at the time. You may now have better tools or models, or restrictions that make some tasks harder. Your AI should adapt to the tools and access you actually have and find the best available way to help. Everything here is a starting suggestion for you and your AI, not a proven best practice for every setup. Feel free to rewrite any of it to better meet your needs.
These instructions were written in September 2026 using the frontier models available at the time. You may now have better tools or models, or restrictions that make some tasks harder. Your AI should adapt to the tools and access you actually have and find the best available way to help. Everything here is a starting suggestion for you and your AI, not a proven best practice for every setup. Feel free to rewrite any of it to better meet your needs.