# First, then

by  
Irene Sandler  
July 31, 2024

> “You CAN have it all. You just can't have it all at once.”  
> — _Oprah Winfrey_

Once upon a time, a young Athenian prince named Theseus volunteered to join a posse of tributes to King Minos of Crete, who would feed them to his pet beast, the Minotaur, in a vast labyrinth. With the help of King Minos’ daughter, he slew the Minotaur and escaped with the other tributes back to Athens on a ship.

To commemorate the feat, the Athenians maintained the ship by replacing pieces as the original wood rotted. Enter the Ship of Theseus Paradox: if the pieces of the ship are gradually replaced, is it the same ship? (Marvel’s [Vision & White Vision explain](https://youtu.be/ldoh71uNZmk?feature=shared)).

Now, imagine the ship as a critical system that manifests the way your company operates, the embodiment of your competitive advantage. How do you maintain the integrity and continuity of the business operation _and_ transform it into something that’s not only seaworthy, but can fly?

Many legacy modernization approaches over the past 20 years do it all in one go: swap out the ship’s wheel with an airplane yoke, or remove the masts straightaway. But this approach risks compromising the functioning of the entire system: you end up with a land-bound chimera that can’t move at all.

So why does this failure mode persist?

Partly, because we’re creatures of [habit](https://www.mechanical-orchard.com/insights/manufactured-certainty). Or we can’t [resist the marshmallow now](https://en.wikipedia.org/wiki/Stanford_marshmallow_experiment), even if it means missing out on more marshmallows later (heck, we might not even be in the same role later). We’re rewarded for “quick wins” over meaningful change (see [Exhibit 3](https://www.mckinsey.com/capabilities/people-and-organizational-performance/our-insights/successfully-transitioning-to-new-leadership-roles?cid=other-eml-alt-mip-mck-oth-1805&hlkid=28cb5657cab3435399c3a9890405bf00&hctky=3017137&hdpid=69365df3-2fad-4ddc-bfd3-ab480078110a)); heroic saves over ounces of prevention.

But these are not binaries. It’s possible to maintain course, clean the slate, _and_ show proof early and often:

- Create a side-by-side replica of the behavior of the system workflow by workflow
- Ensure that each replicated workflow is operating identically to the original system using the same data flows and under the same conditions
- Switch off the original workflow once it’s equivalent

Et voilà! Working software in production! Continuously accelerating progress! Minimal risk! Have your marshmallow now _and_ eat more later!

Rebuilding the ship plank by plank, before introducing “innovations,” ensures that these critical systems are on a stable foundation that’s easy to understand, change, and pivot when the winds of competitive advantage shift.

Now _that’s_ smooth sailing ahead.

### News and Views

Cyberattacks get the headlines, but root causes are less glamorous. _The Economist_ relates the [most recent set of NHS woes](https://www.economist.com/britain/2024/06/13/britains-nhs-reels-from-a-ransomware-attack), noting “owing to crumbling IT services and underinvestment, many trusts lack the capacity to perform even ‘simple upgrades’.” _Ars Technica_ reports on [consolidation risk + cyberattacks](https://arstechnica.com/information-technology/2024/06/cyberattacks-have-forced-thousands-of-car-dealerships-to-paper-for-a-second-day/) at CDK Global.

[Startups are taking longer to get from Series A to Series B](https://news.crunchbase.com/venture/startups-raise-series-a-series-b/), as pointed out by Crunchbase last year. Emergence Capital’s inaugural [“Beyond Benchmarks” report](https://docsend.com/view/pi629vaxy57u9qvb) concurs, also noting that native GenAI companies get an 80% boost for Series A and a 260% boost for Series B compared to AI-less companies.

Well, you know it’s serious when real money is spent. Like [Accenture’s $90 million worth of AI-related bookings](https://www.thestack.technology/accenture-generative-ai-bookings/). But wot’s this? “Clients are finding it hard to scale AI projects.” 🫠

For an entertaining and deliciously profane rant that, ah, creatively picks apart the corporate press release buzzword bingo of AI’s promise, read [this](https://ludic.mataroa.blog/blog/i-will-fucking-piledrive-you-if-you-mention-ai-again/).

### From the Orchard

Incremental modernization or big-bang/all-in-one modernization? In his [latest blog](https://www.mechanical-orchard.com/insights/incremental-vs-all-at-once-you-dont-get-to-choose), Kent Beck offers a nested view.

Tradition doesn't always pave the way for future success; often, it anchors you and your business in the past. Matt Wynne works on [unraveling a different kind of legacy](https://www.mechanical-orchard.com/insights/the-iceberg-model-towards-unraveling-our-patriarchal-legacy), writing about how to use the Iceberg Model to dismantle patriarchal structures.

We hosted the latest of our Mechanical Orchard dinner series in New York some weeks back, and came back with plenty of interesting nuggets. Highlights on [LinkedIn](https://www.linkedin.com/feed/update/urn:li:activity:7212122511897968640).

Companies like ours are [still betting on San Francisco](https://www.bizjournals.com/sanfrancisco/news/2024/06/13/ai-downtown-reset-san-francisco.html)—and we’re in good company.

If you’re keen on joining us for the next dinner, keep an eye out for upcoming events [here](https://events.mechanical-orchard.com/).

Curious to learn more? Say [hello@mechanical-orchard.com](mailto:hello@mechanical-orchard.com).
