What stage in the product lifecycle is best for a product launch? What effort would you put into launching a product that is still very much an MVP?

At the highest level, I love announcing the same product multiple times (shhh: don't tell my PR person). The frequency of message makes it easier for your story to stick. Plus, depending on your customer purchasing cycle in B2B, they often need several months to go through their testing and budgeting cycles to adopt new products. 

The best launches I've done have been by:

- announcing that the product is coming - opening up selected preview by request 

- announcing that the product is in beta - maybe add pricing info or some new detail

- announcing that the product is generally available - adding customer stories

Having said that, it's critical to manage the reputation & competitive risk. You can't say that the product is coming and then it takes 3+ months for the product to go in beta. So, only use the drumbeat approach when you have good confidence and an understanding with your product team. If you don't have that confidence, then stick to announcing at Beta.

This is such a great question!

At some point every PMM will find themselves in a situation when they are asked to launch a product that just isn’t ready. It’s always a mistake to go too early (launching too late isn’t good either but less dangerous).

If it’s truly an MVP and you launch it with the expectation it isn’t people will be disappointed. I think it’s possible to change the narrative around your product but why take the risk? A launch should be a celebration and underwhelming people at the jump isn’t going to get you anywhere. People usually understand a new product won’t be 100% built out, but it should fulfill the core promise you make if your messaging. 

The best thing we’ve done here to help make this easy (because product and leadership WILL ask you to do this even though it’s a bad idea) is to document launches that went to early and why that was bad. And to have some sort of SLA with product, like you can’t enter beta until at least 10 real customers are using the tool, and you can’t launch a product until your NPS hits a certain number. Those rules will save you from launching a product too early and disappointing your audience.

Lets define some terms

Release - product is ready to to be used by customers. Release can be in various forms - limited preview/private preview, beta release, Generall available

Launch - a set of marketing activities to get the word out about a product - to drive usage.

Announce - a simple announcement that a product is likely to be released/launched in the future

In my experience, a product should be launched close to being released. Could be launched as beta, or limited availability (CTA-sign-up for this limited preview), or GA (CTA - use now) . Big fan of doing it as a drumbeat (e.g. launched in a monthly newsletter, and reinforced when it is fully GA in a quarterly event- often with real customer stories accompanying it) and then reinforced at your annual user conference (Check out this product we launched a few months ago - here are some momentum stats behind it). 

Don't keep more than 3 months gap between limited availability and GA. 

In some rare cases, we have announced products several quarters in advance - when the product was technically still in MVP. This is a way to communcate vision and show where we are going /art of the possible for major , market changing products. The announcement is often accompanied by intricate storytelling / demos that demonstrate the art of the possible (not just the MVP)

But such tactics should be used sparingly. You cannot pre-announce everything. You have to take select elements of your roadmap - that are truly market moving, and where you gain competitive advantage by pre-announcing them (e.g. give your sales reps more talking points about roadmap and vision in highly competitive, often multi-year contracts). And then follow through with the real product in a reasonable time-frame (6 months). Any of these could become failure points and destory credibility. 

You may wish to consider a launch just for existing customers, while it's still an MVP.

This will give you a chance to fine-tune the launch messaging, garner feedback (and quotes) and prepare you for a full external launch (which could include an announcement, public beta, general availability etc.)

Announce and launch are two different things. You can launch without announcing and you can announce without launching. Launch is a process (which Manav is alluding to), it's not a point in time.