Stephan Schmidt - January 22, 2025

10 Rules to Reduce Time to Market

A sweet and bitter truth for me as an engineering manager and CTO: Time to market can’t be reduced by just looking at development. You can squeeze out 10 to 20 percent by optimizing development, but you can’t get to the next level unless the development department is really bad or totally blocked by technical debt.

The lever to get things faster, to reduce time to market, lies somewhere else: in decision making and fast and often releases. Optimize the whole way from idea to a feature generating money to reduce time to market.

And while this is true, in every company I’ve looked over the last twenty years, everyone is just focusing on faster development.

Executives are especially bad. They take a long time for decisions, then push development hard.

Quality of decisions for me, speed of execution for you.

Which means I take my time because I’m important, you just do your work as fast as possible, you’re decisions are not important. What a culture.

What Is Time To Market?

Time to market is the time it takes from an idea to a feature generating money. The time from when someone has an idea in the morning to the feature being released and used - and in the best case generating revenue.

Most people measure this metric wrong. They start from “when development began” and end at “when development ended.” That’s not time to market. That’s just development time. Or they measure time when an item was put into backlog to the item making money. Also wrong.

The right starting point is when someone has the idea. Create a ticket as soon as there is an idea. Don’t talk about an idea if there is no placeholder ticket to capture the time the idea was born. The right endpoint is when someone uses the feature and in the best case, makes money (another reason to release more products and less features, a product being something that you can sell, a feature being something that you already sold with something else). Measure feature usage and celebrate the point in time, when the feature was born, when it was used by a customer for the very first time.

Time To Market Over Time

In the beginning of a company, as a startup, time to market is short. Founders are either developing themselves (prompting themselves :-) or are sitting right next to engineers. Founders are product managers themselves and make the founder <-> PM conversations right in their head.

Over time this changes and time to market gets longer and longer. When founders no longer focus on product, when processes are introduced, more integrations happen, compliance is added, more people are hired, and code gets unmaintainable with complexity - time to market takes a hit - and founders wonder why: “In the beginning we were much faster!” is their desperate cry.

Your software is an iceberg. Things above the waterline and things below. Over time the parts below the waterline get larger and larger. GDPR, SOX compliance, integrations, audit trails, marketing integrations, multi-cloud. All planning takes longer, you need to talk to more people to make something happen. Legal, compliance officers, key account managers, brand managers, cross border project managers. They all want to have a say. All development takes longer because you need to change more things when adding new features and because new features work in unpredictable ways with old features (aka bugs).

The 10 Rules

Faster time to market is a huge revenue driver. Products that are earlier in the hands of customers generate money earlier. If a feature takes one month time to market vs six months, in the first case it already makes money for five months!

But what can you do to reduce time to market and get things out faster?

1. Reduce rework Low bug counts, high quality in requirements means less rework. Less rework means developers are not kept fixing features that were already done - this lowers time to market for that feature and the next where developers can start earlier. because they are not blocked by rework.

2. Frequent releases with continuous deployments Features developed but not released do not bring in money. Feature flags are great, but fear to turn them on.

3. Stop working in parallel - everywhere Focusing on one thing to get through development speeds this thing up. Everyone focus on this one new product instead of the company working on ten initiatives in parallel.

4. Smaller features Less to code, faster done, lower time to market. Don’t create this killer features, don’t fall in the trap of “Only if we add this and this and this the customer will use it.” Don’t fall in the trap of throwing more and more on a moving train until it comes to a standstill.

5. Quicker decisions If decisions take a long time they block development and slow down time to market. This is the biggest one. While everyone focuses on speeding up development, no one focuses on decisions. Wherever I looked, decisions were soooooooo slow. Fear everywhere to make decisions. Founders and CEOs without vision: Then it’s hard to make the right decisions.

6. Rigorously cancel meetings Meetings not only waste time that could be spent otherwise, they also let people “wait for the next meeting” instead of doing something now. Double so with remote. Or not, in every meeting I have been over the last month with >5 people the majority turned off the camera and worked something.

7. Shorten product management cycle Things that get faster out of product management get faster into development and faster to market. Have a process to hardening (AI-)prototypes so they can be put directly in production.

8. No single point of failure or bottleneck Bottlenecks increase time to market because work waits in line to be done. Single points of failure - when failing - block everything and throw time to market into chaos. Control is your biggest bottleneck.

9. Don’t look at code If everyone needs to read AI generated code, if every PR needs to be checked by hand, they pile up, people are drained and features take a much longer time to reach customers (but check your compliance officer!)

10. Use AI everywhere Everywhere. Time to market is not only in development, but design, customer understanding, making decisions, requirement engineering - everywhere. If you use AI everywhere time to market plummets.

The Biggest Driver: Buffers

One big driver for time to market is the buffer between phases. People come up with ideas, they go in a roadmap. Then there is a decision from the CEO or board. Then they go to a backlog and rest for some days. People plan ideas into features, then they rest for some days. Developers develop the feature, then it again might take days through QA and release. Then there is no product marketing, so the feature is not used for a long time (at the level anticipated - longer time to market!)

These buffers, when work is ready for the next phase but is waiting in a buffer, add up to a lot of time to market.

The Second Driver: Control

We thought software development was about throughput, or about building the right thing for the customer. It wasn’t. And it isn’t. Software development is about control (See my theory of control ). Control gates (meetings, loops, etc.) are put in place so the CEO, CPO, CTO are confident the right things are build - what they think is the right thing (features according to strategy and vision and brand and positioning and key customer wishes). But these control loops break the flow, loop back, stop work, lead to rework and are big bumpers for a faster time to market.

Reduce control as much as possible in the age of AI and you not only will reap the biggest gains from AI usage but also reduce time to market.

Seven Hypotheses

1. Measure or you don’t know what is going on [1]
2. If a feature is worth developing, it’s worth going to market sooner.
3. Time to market is a huge lever.
4. Buffers are the biggest driver for longer time to market.
5. Everyone can make faster decisions.
6. More than 50% of time is spent before development starts.
7. Control increases time to market.

[1]: I suggest: idea -> development start, development start -> feature being used.

PLUS time to market in software development is a solved problem. There is no excuse not to follow best practices - even when people struggle on how to properly adapt AI. The challenge is everything that happens before development starts and after development finished.