OKR Generation from Client Data — AI-Assisted Drafting

Overview

A recurring challenge in account management is the blank-canvas problem: when it's time to set OKRs for a client, account managers often don't know where to start. Mark built an internal tool that addresses this by automatically generating a first-draft set of OKRs for each client, grounded in real performance data. The output isn't meant to be final — it's meant to give account managers something concrete to react to and refine.

How the Tool Works

The tool pulls together everything known about a client and uses it to propose objectives and key results:

Client context inputs:
- Business type (B2B vs. B2C)
- Commerce model (e-commerce, local service, regional, national)
- Industry and service mix

Performance data inputs:
- Google Ads metrics
- Google Search Console data (traffic, impressions, click-through rates)
- Current conversion rates
- Any other available client analytics

From these inputs, the tool generates OKRs that are specific and numeric — e.g., "Traffic is currently X, and needs to reach Y to hit the conversion target." The objectives are tied to the actual levers available for that client's business model.

Workflow Integration

The intended process is:

Tool runs against client data and produces a draft OKR set
Draft is sent to the responsible Account Manager
Account Manager reviews and adjusts — "not exactly that, more of this, less of that"
Refined OKRs are used for the client's planning cycle

This replaces the prior approach where account managers were expected to generate OKRs from scratch, which frequently resulted in either blank canvases or generic objectives disconnected from actual client metrics.

"If you tried to do all the stuff the tool is doing, it would take you days to go to all those places and look up all those numbers and try to figure it out." — Mark Hope, Feb 20 2026

Why This Matters

Reduces cognitive load on account managers at the start of each planning cycle
Grounds OKRs in data rather than intuition or guesswork
Scales consistently across all clients without requiring each AM to be a data analyst
Creates a feedback loop — AMs correct the draft, which surfaces where the tool's assumptions are off

Limitations and Caveats

The generated OKRs are explicitly a starting point, not a finished product
Account managers must still apply client relationship knowledge and strategic judgment
Tool quality depends on data availability — clients with sparse analytics will produce weaker drafts
Requires AMs to actually engage with and refine the output rather than accepting it passively

^[1] — ClickUp is the system where OKRs and tasks ultimately live
^[2] — The three-AM structure this tool supports
^[3] — Related internal tooling built on client data