Learn

How to keep a running summary of each client relationship


Every agency and consultancy runs on relationships. The quality of your work depends on knowing what was promised, what was delivered, what changed, and what matters to each client right now. Yet most of that knowledge lives in fragments: a Slack thread from three months ago, a Google Doc someone forgot to update, a meeting recording nobody rewatched, an email chain buried under hundreds of others.

The typical response is a CRM. But CRMs only know what people type into them, and client-facing teams are busy doing the work, not logging it. Fields go stale. Notes trail off after the first few weeks. By the time someone needs the full picture of a relationship, they are stitching it together from half a dozen sources and the memories of whoever happened to be on the last call.

Fabric takes a different approach. Its self-writing docs use a client relationship tracker output type that creates and continuously updates a per-client page drawing from every connected source. No one has to fill anything in. The summary writes itself because it is synthesised from the conversations, files, meetings, and decisions that already exist across your tools.


What a client relationship summary looks like in Fabric

A client relationship page in Fabric is not a static profile or a list of fields. It is a living document, powered by Exabase M-1, that weaves together deal history, meeting summaries, key decisions, deliverables and their status, and open items. Think of it as the briefing you would write if you had unlimited time and perfect recall of every interaction across every channel.

When you connect your tools -- Slack, Gmail, Google Drive, Notion, calendar apps, and more from over 50 available integrations -- Fabric's cross-source synthesis engine begins linking related information. An email thread about revised deliverables gets connected to the meeting where the scope change was discussed, which in turn links to the updated brief sitting in Drive. The client relationship tracker pulls all of this together into a single coherent narrative for each client.

This is not simple aggregation. Fabric's memory engine builds persistent context about each relationship over time. It understands that "the Q3 campaign" mentioned in a July Slack message, an August meeting transcript, and a September email attachment all refer to the same body of work. That contextual understanding is what turns a pile of data into a useful summary.


Setting it up

Getting started requires very little. Connect the sources where your client communication and work product live, and Fabric begins building relationship pages. There is no schema to design, no fields to map, no import to run. The system reads what is already there and structures it.

For agencies managing a portfolio of clients, this means every account gets its own continuously updated page from the moment the connection is live. For consultancies with longer, deeper engagements, the pages grow richer over time as more context accumulates.

You can shape what each summary emphasises. If your team cares most about open action items and upcoming deadlines, you can guide the output type in that direction. If executive stakeholder sentiment and decision history matter more, the tracker can foreground those. The structure adapts to how your team works rather than imposing a rigid template.


How it stays current without anyone updating it

The central problem with every client-tracking system is maintenance. People start strong, then stop entering data as soon as the work gets busy. Fabric sidesteps this entirely because the data entry is the work itself. Every email you send, every meeting you attend, every document you share is already flowing through your connected tools. Fabric watches those streams and updates each client page accordingly.

After a client call, the meeting summary, action items, and any decisions get folded into the relationship page within minutes. When a deliverable ships and the client responds over email, that response and its sentiment become part of the record. When a teammate has a side conversation in Slack about timeline changes, the tracker picks it up and adjusts the open items.

This is where Fabric's AI knowledge base for clients becomes indispensable. It is not just storing information; it is maintaining a structured, queryable understanding of each relationship that any team member can access at any time.


Preparing for client interactions

One of the most immediate benefits is what happens in the five minutes before a client call. Instead of scrambling through old emails or asking a colleague what was discussed last time, you open the client's page and read a current summary. Every recent interaction, every outstanding item, every piece of context you might need is already there.

Fabric Tag makes this even faster. You can tag client-related content as you encounter it, and that tagged material feeds directly into the relationship summary. If a colleague flags a tricky email exchange or an important piece of feedback, tagging it ensures it surfaces in the right client page rather than getting lost in the flow.

Before a quarterly business review, you have a complete record of what was delivered, what changed, and what the client said about it along the way. Before a new business pitch to an existing client, you can review the full arc of the relationship, from initial engagement through every project, to speak with the kind of institutional memory that builds trust.


Finding anything across a client's history

Relationship summaries are useful for the big picture, but sometimes you need a specific detail. What did the client say about the logo options in March? When did we agree to extend the retainer? Who on their side approved the revised budget?

Fabric's semantic search lets you ask these questions in plain language and get answers drawn from across every connected source. You are not searching a single app or folder. You are searching the full context of the relationship, including emails, meeting transcripts, shared documents, and Slack conversations, all at once.

This is a meaningful shift from the way most teams operate. Normally, finding a specific detail means remembering which tool it was in, then searching within that tool, then hoping the right keywords surface the right result. With Fabric, you describe what you are looking for and the system finds it, regardless of where it originated.


Giving AI agents scoped access to client context

Modern agencies and consultancies increasingly use AI to draft proposals, summarise research, prepare briefing documents, and handle other knowledge-intensive tasks. The challenge is giving those AI tools enough context to do useful work without exposing your entire knowledge base.

Fabric's agent access solves this with scoped, time-limited permissions. You can grant an AI agent access to a specific client's materials for a specific task and duration. The agent can read the relationship summary, search relevant files, and use that context to produce better output, all without persistent access to anything beyond what it needs.

Through MCP and agent access, you can also use Fabric from within ChatGPT or Claude directly. Rather than copying and pasting context into a chat window, your AI assistant connects to Fabric and pulls the relevant client information itself. This keeps your workflow fluid and your client data secure behind Fabric's AES-256 encryption and CASA Tier 2 compliance.


Sharing client-facing content

Not everything in a client relationship stays internal. Agencies and consultancies regularly need to share deliverables, reports, and updates with clients. Fabric's published spaces let you share selected content through a clean, branded experience.

What makes this particularly useful alongside relationship tracking is link analytics. When you share a proposal or a report, you can see whether the client opened it, how long they spent with it, and which sections they focused on. That engagement data feeds back into your understanding of the relationship, so your next conversation can be informed by what the client has and has not reviewed.

For teams looking to create a more structured sharing experience, Fabric supports full client portals where clients can access their materials in one place, reducing the back-and-forth of email attachments and shared drive links.


Scaling across a portfolio of clients

Individual client knowledge is valuable. Portfolio-wide visibility is transformative. When every client relationship has a maintained summary, agency leaders and practice heads can see across the entire book of business without scheduling update meetings or reading through project management tools.

Which clients are approaching renewal? Where are open issues accumulating? Which accounts have gone quiet? These questions become answerable at a glance rather than requiring a round of status emails.

For agency owners in particular, this kind of visibility changes how you run the business. You can spot at-risk relationships before they become problems. You can identify upsell opportunities based on the full history of what a client has engaged with. You can onboard new team members onto an account in hours rather than weeks, because the institutional memory is written down and current.

Teams that want to go deeper into organising their client materials can pair the relationship tracker with structured client file organisation, giving every account both a narrative summary and a well-ordered library of related assets.


Frequently asked questions

Do I need to set up a CRM integration for this to work?

No. Fabric's client relationship tracker does not sit on top of a CRM. It builds relationship summaries directly from your communication and collaboration tools: email, messaging, meetings, file storage, and project management apps. If you use a CRM and want to keep it, that is fine, but Fabric does not depend on one. The point is to create a summary that writes itself from the work you are already doing.

How does Fabric decide what belongs to which client?

Fabric's memory engine and cross-source synthesis identify client-related content by recognising names, email domains, project references, and conversational context across your connected tools. Over time, it builds an increasingly accurate map of which conversations, documents, and meetings relate to which client. You can also use Fabric Tag to manually associate content with a client when needed.

What happens when multiple team members work with the same client?

The relationship page synthesises interactions from everyone on your team who has connected their tools. If one person handles strategy calls and another manages day-to-day execution, both sets of interactions feed into the same client page. This gives any team member a complete view regardless of their individual involvement.

Can I control what information appears in a client summary?

Yes. You can guide the output type to emphasise certain categories of information, such as decisions, deliverables, or open items. You can also exclude specific sources or channels if they contain irrelevant noise. The summary adapts to your preferences rather than dumping everything in without discrimination.

Is client data kept separate and secure?

Fabric uses AES-256 encryption and holds CASA Tier 2 certification. Client data remains within your workspace, and agent access is scoped and time-limited, so AI tools only see what you explicitly permit for the duration you specify. There is no persistent broad access to your knowledge base.

How quickly does the summary update after a meeting or email?

Updates happen within minutes of new content appearing in your connected tools. After a meeting ends and the transcript is available, the client page reflects the key points, decisions, and action items from that conversation shortly afterwards. The same applies to emails, Slack messages, and document changes.

Can I use this for prospective clients as well as existing ones?

Yes. The relationship tracker works for any ongoing relationship, including prospects in an active sales process. As you exchange emails, have discovery calls, and share proposals, Fabric builds a page that tracks the progression of the opportunity. This is especially useful for managing multiple clients at different stages simultaneously.

What if I want to share the client summary with the client themselves?

You can publish selected portions of a client page through Fabric's published spaces. This is useful for creating transparency around project status, deliverables, and open items. Link analytics let you see whether the client engages with the shared content.

Does this replace project management tools?

Not directly. Fabric's client relationship tracker focuses on the narrative arc of a relationship: what has happened, what was decided, what is outstanding, and how the engagement is progressing. Project management tools handle task assignment, sprints, and workflow automation. The two complement each other, with Fabric providing the contextual layer that project management tools typically lack.

How does Fabric handle clients who communicate across many channels?

This is where Fabric's cross-source synthesis is most valuable. A client who emails about budgets, discusses creative in Slack, gives feedback in Google Docs comments, and makes decisions on video calls generates a fragmented trail that is nearly impossible to follow manually. Fabric connects all of these threads into a single coherent summary, regardless of channel.


Related pages


The workspace that thinks with you.

Ready when you are.

The workspace that thinks with you.

Ready when you are.

The workspace that thinks with you.

Ready when you are.