Week 2: Current State and North Star
Week 2: Current State and North Star
Before I start running experiments I want a frame to evaluate them against — otherwise I’m just collecting anecdotes.
So Week 2 is the strategic pit stop. An honest read of where AI already sits in my workflow today, where it sits at the org level, and the single sentence I want to hold every future experiment up against.
Current state — what I’m actually using
Today’s AI footprint in my personal workflow is narrower than I’d like to admit:
- A notetaker on most customer calls
- A couple of sales prospecting and research tools, mostly for tech-stack discovery on accounts before discovery calls
- Ad-hoc chat assistants for the usual: drafting, summarizing, rewriting
What’s missing is more interesting than what’s there:
- Nothing is keeping me organized across accounts
- Every tool lives in its own tab and demands its own context load
- Context-switching is still the largest invisible tax on my week
Current state — what the org has shipped
Zoom out one level. At the org level, the AI tooling we’ve rolled out skews three ways:
- Sales-team-first: built for AE outcomes, with the SE workflow bolted on as an afterthought
- Out-of-band: lives outside the surfaces I actually work in (CRM, email, Slack, the product) so adoption decays fast
- Slop-prone: confident-sounding output that needs a real SE to catch the wrong thing before it ships
The frustrating part is that all the building blocks exist. The models work. The data is reachable. What’s missing is a clear opinion about what we want any of it to do.
Zooming out before zooming in
The temptation right now is to grab the next shiny agent and run the experiment. I’ve watched myself do that for a year. Net result: a graveyard of half-evaluated tools and zero coherent story about which ones moved the number.
If I’m going to spend the next twelve months publicly transforming my workflow, I need a frame that tells me which experiments matter and which ones are just novelty. That frame has to start from how this job is actually scored, not from how AI vendors describe themselves.
The north star
The honest version, stated plainly:
Use AI to close more business, and bigger business, faster — so the quota number gets hit, the commission check gets bigger, and the customer gets to outcomes sooner than they would have without me.
From the SE seat specifically: get to technical validation that’s demonstrably tied to a business outcome, faster than I could without AI in the loop.
Efficiency vs. effectiveness — the evaluation principle
Every AI improvement in this series will get scored on one question: did it move me toward the north star?
That splits into two flavors:
- Efficiency — same outcome, less time. Time recovered should flow into shaping work that moves deals.
- Effectiveness — better outcome at the same time spend. Sharper technical validation, tighter business framing, fewer things missed.
Either is fine. Both are great. Neither counts if the recovered hours go into more pattern-matching instead of more shaping.
What’s coming next
Week 3 is when the experimentation begins: an agent drafts a discovery brief for me. Now I have something to score it against — not “was the brief good?” but “did this move me, even a little, toward closing more business faster?”
A question worth sitting with if you’re an SE: what’s your north star — the one sentence that tells you whether the next AI tool in your stack is worth your time?