How to Launch New Product Features

Intro

I originally wrote this short piece for a few colleagues to convey my personal experience and thought process for my personal brand of product feature discovery & validation but thought posting it on Medium would be the more utilitarian albeit self-serving approach ;)

I draw on my experiences in tech, finance & e-commerce; working in a multitude of roles (engineer, project manager, product manager) that have focused on product launches (new equity funds, new websites, new enterprise tools). The successful ones have a set of commonalities while the not-so-successful ones were missing a few key components. I’ll only lay out the components in successful launches in this article.

The scope of this piece will be on a current e-commerce website that already has users but is looking for growth from a larger corporate organization.

The Abstract

How can you be certain that a new feature or improvement that you plan to roll out for your product is going to be a hit? Are you using your gut feeling based on personal experience or conducting market research to look for direct cues from the customer on what to implement?

There’s a dichotomy for discovering new product features: Heart & Mind. The heart represents the intuitive side of decision-making while the mind represents the data-driven approach to decision-making.

Example:

[Heart] Derive the desires of your customer from your observations or personal anecdotes.

then

[Mind] Verify those desires from surveys or product usage data.

On the flip side

[Mind] Use statistical methods like Chi-Square Test of Independence to find causation between two different data sets of the same users.

then

[Heart] Conduct 1-on-1 user interviews on hand-picked data points from your quantitative analysis to verify your data correlation assumptions.

The caveat to the Heart & Mind is they can be used interchangeably in the process but usually never alone. Think of it as, you can’t have the yin without the yang. Balance, Daniel-san!

Trust the Process

All the product development articles & classes you’ll find out there can be distilled back to the Scientific Method. The nitty gritty procedures may vary depending on your industry but is for the most part applicable at every stage.

My past experience in undergrad engineering really hammered in this methodology and I would highly recommend the same to anyone in the field of Product as it has proven successful for me in the past. The process allows for objective and healthy debate in an organization of any size which allows the good ideas to rise to the top. It has the added bonus of usually being the most risk adverse (my favorite).

You’ll find many articles out there that will dive deep into any one or many of these areas but the distillation will usually be the same.

Okay, how do I apply this?

I have a 10-step process that was derived from the scientific method and my past experience to basically launch a new feature on an e-commerce site. You really want to spend most of your time on step 1–4 and really get that right. The more you iterate on that piece and really understand the underlying assumptions and validations of those assumptions, the better chance you have to launch a new product/feature that will be successful!

1. Vision — What is your mission? What problem do you want to solve? What is your value proposition? Why is it so important that we provide the product that we do? Why do customers care about this particular solution to their problem? This can also be an assumption you’re making about a particular set of customers that you’re looking to validate.

2. Market Research — Who are your customers? What are their behaviors? I’ve found that surveys and site usage data create a nice view of not just how people are using your site but what they’re thinking as well. Create a persona or a few to really personify your largest user bases.

3. Strategy — How will you go about achieving your mission? How will you test your hypothesis? If you’re running an e-commerce site, are you going to add a new feature? If you’re in the consumer electronics industry, are you going to offer a new earphone product?

4. Business Plan — What KPI’s do you need to hit for your new feature on your e-commerce site? What kind of support will your users of the new feature need? The more detail the better. Include your roadmap and timelines here as well. What are the milestones you intend to hit? (ie PRD, Demo POC, Growth plan, expected cost, expected revenue). What they don’t tell you is there’s a lot of negotiating with various product/engineering teams to get the features you want. Rarely can you deliver a new feature with just one engineering team’s help. You will need to be great at convincing and influencing teams.

5. Release Plan (Sprinting & Iterating) — This is usually where the Scrum/Agile process comes into play. You’ll need to plan out your execution plan with your Engineering/UX/QE team and figure out technically which components you’ll need to build to implement according to your timelines. Included but not often discussed is A/B testing & stress testing (especially needed for large scale e-commerce sites). Spend as much time as you can with your teams to make sure everyone understands the plan. You need to spend time with engineering, design, QE, business units, marketing. Also, great user story writing is so so so important! Communication with your key stakeholders is also key to protecting your team from unneeded corporate political jabs from within that will hinder progress.

6. Marketing — You’ll need to work with your marketing team to target your key demographics and get the word out about your new feature. Email campaigns, TV advertising, social media advertising are the most straight forward ways of getting the word out (some more costly than others). Don’t forget old-fashioned channels such as paper flyers in key physical locations where your users are lurking!

7. Launch Day Pt. 1 — Leak 1-5% traffic — All hands on deck! You never want to just flip the switch on a large scale website and flood your servers with new connections it might be not able to serve up. A slow turn of the valve is recommended so that it’s easy to shut off the valve preemptively if needed.

8. Fix bugs! — There’s always unforeseen bugs to fix. Add them to the backlog and prioritize to fix asap!

9. Launch Day Pt. 2 — Slow leak from 5-100% traffic

10. Fix bugs! — There’s always more bugs to fix, but you’ll want to focus on the ones that will stabilize your production environment.

Outro

Parts of the process have been distilled but you’ll be able to find a TON of material on each one of those points online. The key takeaways are:

1. Product/Feature launches are a disciplined balance between Heart & Mind
2. Trust the Scientific Method
3. Apply the Scientific Method to your field of interest
4. Bring the passion & enjoy your work!