Solution State Model (SSM) - a Case Study ["TECH Workflow"]

•

Ever faced situations that could get so chaotic, you felt like quitting is the best & only solution? I have & I wished I had the clarity then, would've made a world of difference. But don't you worry! Over this thread 🧵 I'll prepare you to wade through such situations... 1/n

Yes. Apart from the visual appeal mental models could have a great bearing on the person viewing it as it could get etched in memory. I published the "Solution State Model - SSM" recently & it was an instant connect with all my peers. Link: mgmtinc.substack.com/p/solution-state-modeling-ssm 2/n

Now extending that thought over my threads I usually dive into a CASE STUDY where I: - pick a situation - define the problem - decipher it - build an understanding - apply the solution / mental model as discussed Nothing gets more pragmatic, does it? So, let's dive in. 3/n

CASE STUDY: A legacy payments app that's LIVE across 100+ nations has massive interfaces depending on & feeding off it. Now, some users dial in to complain about these: - desktop app seems stuck - web app hangs - mobile app crashes, closes down abruptly 4/n

The no. of affected users seem to be less based on the dial-ins. But in reality, that could be noncorrelative to the magnitude of the problem, as it's just in, something yet to come to light. At this stage, it could still feel like it is MANAGEABLE (refer canvas). 5/n

A probe is ordered & the tech support team gets down to work over assessing the situation & let's say their first-level findings are: - most payments aren't reaching the gateway - a delay is seen - too many timeouts which isn't manageable anymore, so canvas now changes to: 6/n

Working backwards & get into RCA (Root Cause Analysis): Starting from (3) - Timeouts are a way to stop errors & maintain homogeneity. If outcome isn't reached within a stipulated time it is purged which is linked to Delays (2) FYI - here's a typical architecture of a queue 7/n

Payments not reaching Gateway (1) could be linked the other two points. RCA points to internal errors & leads to these findings: - queue system is non-responsive (back) - queue seems to hold full capacity - dispatch service is broken (front) Canvas points to CHAOS now: 8/n

This is a good point for teams to delve deeper into each problem leading to say, these findings: - implement intelligent dispatch tying it to the receiver - dispatch service hung, needs a refresh - check dispatch efficiency MAJOR CAUSE - queue seems architecturally weak 9/n

- implement auto-heal & regulate queue capacity based on load - check queue buffer size & increase it if needed - reciprocate with increasing capacity at the receiver end - deep test the front & back of the queue With all that clarity, our updated canvas is here: 10/n

But, if you look at the steps listed under CHAOS, you'd see how they're both good to be broken down into further steps over another Sub-states which deserve to be represented on a canvas of it's own owing to the magnitude of the tasks. Like so: 11/n

Here are the descriptive sub-states over individual canvases: SUB STATE 1 - implement intelligent dispatch tying it to the receiver SUB STATE 2 - implement auto-heal & regulate queue capacity based on load SUB STATE 3 - check dispatch efficiency Here are the images: 12/n

The Solution State Model (SSM) could be applied for anything complex from a real life situation to a very deep workflow. When the case study here was TECH, it could go well with something totally different like an elaborate painting or planning complicated & nuanced steps. 13/n

NOTE: Another common complaint is how everything here in the final image when it gets into the representation of SUB-STATES could look so visually complex & overwhelming. And, that's just where "Cascading windows" kind of an UI could come to your rescue... 14/n

That's it folks! If this thread resonated, you'll love my newsletter. Sign-up here: mgmtinc.substack.com/ Also, please do me a favor: 1. RT the first tweet 🔝 marked "1/n" 2. Follow me on @bgpinv for daily product musings Thanks for reading 🙏 Comments are welcome!