Field Notes

Most small businesses buy a CRM when sales feel inconsistent. The intent is good: track leads, follow up, close more.

But in diagnostics, the CRM is often the last step in a broken chain. The chain breaks earlier: no ownership, slow response, unclear handoffs, and no definition of "done."

When fundamentals are missing, a new CRM becomes a very expensive spreadsheet. People enter partial data, skip steps, and default to texting anyway.

A fixer starts with the lead journey: what happens in the first 10 minutes after an inquiry? Who owns it? What's the minimum data needed? When does a lead become "won" or "lost"?

Tune the CRM to the process, not the other way around. Often the best first move is simplifying: remove fields, tighten notifications, install one follow-up cadence.

If you can't say "every lead gets contacted within X minutes," you don't have a CRM problem. You have an ownership problem.

Automation should reduce effort. But in many small businesses, it increases it.

The pattern: tools get connected before "correct" is defined. Errors multiply faster.

The business starts babysitting automations: duplicates, wrong messages, brittle integrations, re-entry work.

Automation is a multiplier. It amplifies whatever is underneath. Messy process → amplified mess.

Before automating: ownership, inputs, and a definition of done. Then automate only deterministic steps.

Add observability: logs, alerts, and a weekly audit so automations stay trustworthy.

If every important decision routes through the owner, the business has an API problem: the owner's attention is the interface.

Delays, stalled leads, and inconsistent delivery follow.

The fix isn't "delegate more" abstractly. It's redesigning decision paths so common cases don't require the owner.

Define 3 tiers: team decides within rules, team proposes/owner approves, owner decides.

Delegation requires shipping context: rules, examples, and escalation criteria.

When decision contracts are clear, the owner stops being a bottleneck and becomes a multiplier.

Most missed revenue isn't lost because you can't generate leads. It's lost because handoffs are sloppy.

The moment after first contact is where things die: a note in someone's head, a sticky, a vague "I'll follow up."

A fixer looks for seams: marketing→sales, sales→scheduling, scheduling→delivery, delivery→billing.

Every seam needs: owner, next action, due time, definition of done.

Tools can enforce seams, but only after the handoff is defined.

Start with one seam (often inbound lead→booked appointment), run it for two weeks, then expand.

Many owners buy software to buy relief. The hope is that a tool will remove chaos.

Tools don't remove chaos. Tools implement decisions. If the decision is unclear, the tool adds friction.

Start with one question: what is the one operational problem that, if fixed, reduces stress and increases profit?

Choose the smallest intervention. Often it's workflow tightening, better intake, or clearer ownership—before new software.

If you do need a tool, buy it with a plan: configure the minimum, ship it, measure it, iterate.

Don't buy platforms you can't maintain.

Many small businesses track vanity metrics: traffic, followers, total calls, total quotes.

Those don't reveal the constraint.

Fixer metrics are closer to work: lead response time, show rate, cycle time, rework rate, throughput, margin per job.

If response time is slow, nothing else matters. If cycle time is long, cash gets trapped. If rework is high, margin disappears.

Start with a weekly scorecard of 5 numbers. Make them observable. Attach owners and actions.

Many SOPs fail because they're long and ignored. They describe what should happen, not what to do when reality happens.

Write SOPs as decisions: "If X, then do Y."

Collect the top 10 questions your team asks. Those questions are your real SOP.

Write a one-page decision tree supported by examples.

Store it where the work happens (templates, CRM notes, ticket fields).

Update when it's used. SOPs are living contracts, not theater.

You don't need a six-month transformation to feel improvement. Two weeks is enough if you focus on one constraint.

Pick one workflow that touches revenue: inbound leads, scheduling, invoicing, fulfillment.

Define handoffs, assign owners, remove steps, then add simple automation: reminders, templates, one dashboard.

Run it with real work. Fix what breaks. Document the final version. Move to the next constraint.

The key is cadence: weekly review, one change, repeat.

Resistance usually isn't laziness. It's cognitive load.

If a system adds steps, duplicates entry, or punishes busy people, it will be bypassed.

Design for relief: fewer fields, fewer clicks, clear defaults, fast paths for common cases.

Incentives matter: if only failures are noticed, people hide. If clean handoffs are rewarded, adoption rises.

Roll out changes in increments. Train using real scenarios. Make it feel like help, not surveillance.

Tool selection hurts when you try to find the "best" tool. The best tool is the one you can operate.

Rule: pick tools you can maintain with the team you have.

That means: ownership, simple billing, strong support, predictable behavior.

Tie requirements to outcomes ("reduce response time," "remove re-entry"), then test 2 tools for 1 week each on real work.

Choose what makes work easier. Commit. Stop shopping.