Week 1: The Audit — Taking Stock of My Current State Before AI

I promised a workflow audit last week and here it is. Last week consisted of eight different SE workflows with no new AI tooling deliberately layered in. The point is an honest “before” picture so we can measure the “after” picture.

Why audit at all

You can’t measure transformation if you don’t measure the starting state. I want to know which tasks ate the most time, where the friction in my day lives, and which tasks I dread vs don’t.

The week at a glance

The eight workflows:

Discovery meetings (x6)
Post sales calls (x3)
Post sales technical questions via email and DM (too many to count)
Technical Account Review meeting
RFP submission
Custom demo build
QBR Prep - data gathering
Operational tasks

1. Discovery meetings (9 hours)

Six discovery calls in 5 days totaling ~4 hours of meeting time. More discovery than the typical week, but sometimes that’s how the cookie crumbles. Prep time for each call took about 30 minutes and another 20 minutes for post-call debrief and follow-ups. In total, 9 hours were spent on discovery related workflows.

Right off the bat, most of the friction was during the prep phase. The pre-call details were very light: company, prospect name and title, and a vague reason for the call. The discovery calls themselves were mostly interesting. I am always curious to understand customer challenges and how they align with the overall corporate business initiatives. Calls are already using AI transcription and notetakers. The post-call debrief and follow-ups felt like they could be more efficient, although debrief calls were very helpful to stay in-sync with the sales reps.

2. Post sales calls (4 hours)

Three hour long post sales calls with ~20 minutes of prep for each. The prep mostly consisted of re-loading all the context from where each customer left off, gathering supporting documentation where necessary, and a minor validation of an open question from the previous meeting. Total time: 4 hours. The biggest challenge here was context-switching, getting back up to speed from where we last left off.

3. Postsales technical questions (4 hours)

Six to ten inbound questions via email from a customer or DM from a rep. About half of these questions are answered from our public docs. Another quarter could be found in our internal docs. The rest required looping in an SME. The friction here is the fragmented work. Each inbound question was “urgent” to someone, whether that is the customer or the rep. Chasing down the correct SME to loop in and follow-ups also added to the friction.

4. Technical Account Review meeting (8 hours)

Situation: An existing customer with a contract renewal in ~90 days. Champion is already identified within the org. Champion is interested in understanding how they can both get the most out of their existing contract and how they can improve their security posture with additional products. A budget has been allocated for this project that will roll into the renewal.

Roughly 7 hours of prep starting with piecing together the customer’s existing product set. Then analyzing logs and their setup, noting down specific areas of improvement. A balance had to be struck in terms of validating their current setup, providing recommendations to get the most value of their existing product set, and delivering recommendations of upsell products to help with their overall security posture. Fortunately, this customer did a good job onboarding their contracted products.

The biggest challenge was building a substantive yet concise deck with room for fluidity in the event any recommendation needed a live drill down/demo. This was by far the highest-leverage work all week. The prep, while long, is what made the meeting matter.

5. RFP submission (6 hours)

Not much to say here. A last minute RFP request and short deadline. Getting answers is simple enough with 3rd party tools that host libraries of previous answers. Some light AI in the background here with these tools. The challenge was reviewing and identifying source material to validate claims. Overall, the work itself is quite tedious. This area seems ripe for optimization with AI and lots of AI RFP tools seem to already exist in the market.

6. Custom demo build (4 hours)

Reading documentation on the different features that would need to be pieced together. The rest of the time was spent setting up the environment and some light troubleshooting. The work itself was rewarding and the time savings of building out a custom demo for the customer over a full blown POV is worth its weight in gold. The biggest friction here is finding the cycles to work on this without context-switching which can multiply the time involved.

7. QBR prep - data gathering (2 hours)

This task was solely focused on data gathering for the QBR. The narrative will come later. The data gathering included pulling usage data from our analytics platform, clicking through the dashboard to note down usage, and reconciling the two when they disagreed. The friction here was the repetitiveness. Repeated data gathering tasks across multiple customers quarterly is a chore.

8. Operational tasks (3+ hours)

At the very least three hours scattered across the week in 10-15 minute chunks. This includes updating the CRM, status notes, internal chat threads, and calendar triage. The operational tasks and pipeline notes feel like death by a thousand cuts. While understanding its necessity, the dread that comes with this task makes it feel at least 2x as long.

Audit exercise retrospective

Highlights

Roughly 40 hours of work across the eight items. Not counting time spent working in my lab environment, technical enablement, etc.
The ratio of presales to postsales work
Dread ranking (high → low): RFP > Operational tasks > QBR data gathering > Postsales questions > Postsales calls > Discovery prep > Custom demo > Technical Account Review.
The time spent ranking does not match the dread ranking. The avoidance gap is real for those tasks in the high dread ranking.

Last week was a full, yet representative week. Most weeks are filled with tasks that are necessary but eat up cycles. A lot of these yield little in terms of closing new business and driving quota attainment. The dread ranking tends to skew very high for postsales engagements.

What’s coming next

Having gone through the workflow audit exercise, it only makes sense to provide a current state of AI adoption within my workflow. Some of it was highlighted above, while other AI tools were not mentioned as they were minimally used. That said, it will help reinforce the starting state.

To help ground the rest of the journey and ensure we are moving in the right direction, I will clearly define our North Star.

A reader question to think about: if you ran this audit on your last work week, what would your dread ranking look like?