Legacy and the audience you hold

"Should I keep the legacy, or build new?"#

A common feeling: the mental load of dealing with legacy code is far bigger than the load of finding new customers. A brand-new product lets you be creative, with no risk of breaking what already works.

Both feelings are valid. The trap is treating this as one choice when it's really two:

Finding customers from zero is the hardest part of any business, and your customer list is the part you've already paid for. Throwing it away to escape the code is the expensive mistake.

How to freeze without abandoning people#

Others have solved this: several big themes split into a "Classic" version (kept alive, few changes) and a new product. 37signals kept old Basecamp versions running but stopped building features for them.

1. Maintenance mode: security and compatibility fixes only. Say so publicly.
2. A clear support policy: fixed scope, slower response times, AI-drafted answers.
3. A bridge: existing customers get a discount or migration path to the new product.
4. Optionally, an end date, announced early.

Decide per product with numbers#

High revenue and low hours: a cash cow, keep it. Low revenue and high hours: freeze or set an end date. Everything else: maintenance mode. The last column counts too. Within 4,000 weeks, energy is a real cost.

Two honest warnings#

1. "Zero hurt" lasts about 18 months. If the new thing works, it becomes legacy too. The lesson isn't "avoid legacy"; it's "design legacy you can say no to": narrow scope, clear support rules, no promise of every feature forever.
2. A blank page feels creative partly because it's free of feedback. No customers means no complaints, but also no proof. Don't let freedom become a place to hide.

The principle: products change, the audience stays#

The builder's own insight during the week:

Things change, but audience segments remain. The relationship, the audience, the brand: that's what stays. We keep changing as the environment changes (marketplaces, AI, whatever). So start from one list, publishers, and keep building around that segment. Always.

A list only works if it's alive: people who hear from you regularly, reply and trust you. A cold list of old buyers is worth little. This is how long-lasting small businesses work: Awesome Motive has sold one WordPress audience many products for 15+ years; vertical software companies stay with one industry for decades while their tech changes underneath; Kevin Kelly's "1,000 true fans".

Is "publishers" worth holding?#

The segment is splitting#

Hold the growing half. They're doing exactly what the builder described for themselves: holding a living relationship with their audience while the tools change.

The query: small independent publishers (1–20 people) who publish regularly under their own name or brand and want to own their readers.

Two consequences of the principle:

1. Hold the segment, not WordPress. WordPress is today's delivery tool, as a marketplace was yesterday's channel. If publishers move platforms in 5 years, the business should follow them.
2. Count how much of the list is actually yours. Marketplaces often don't hand over buyers' emails. The number of customers you can email today is the real asset. If it's small, job one is converting buyers into an owned list.

"Should I go wider? E-commerce is huge."#

Short answer: don't trade the list for a bigger, vaguer one. Move toward the part of your segment that has money.

You don't need a big market: 1,000 customers × $200/year = $200k/year. The question isn't "is it big enough?" but "can I win 1,000 of them?"

E-commerce wins on money and loses on everything else.

The bowling-pin strategy#

Geoffrey Moore: win one segment fully, then expand to the pins next to it, where your reputation carries over.

Small publishers → publishers who sell subscriptions → creators and experts who sell courses and memberships → content-led online stores

Each step keeps most of your access and edge and adds more ability to pay. And price on what you earn them, not what you cost them. A theme is a cost; a revenue engine is an investment.