The 4 steps of Growth Hacking: from Agile Sprints to Growth Sprints

Reduce product development and marketing risks by following each step in the growth sprint: analyze, ideate, prioritize, and test.

(article originally published on revsquare.com) — How do you decide what’s in your backlog, and how do you prioritize it? The Growth Sprint process can help with managing your backlog and reducing the risks of overspending money and energy on pursuing the wrong product development route or/and the wrong marketing effort. There are four main principles and four main steps.

The four main principles are aligned with Lean, Agile and Design Thinking:

1. User-centric
2. Data-driven
3. Iterative
4. Product definition and marketing as one

The Growth loop has four steps:

1. ANALYZE

This is the Analysis step of the process, with four main components:

a) Data Dive — A data analyst does a data dive. The main goals are to:

• Define user groups. Who are the different users? What features do active users truly value? How long do they spend with the application?
• Define conversion funnel for each user type. What is the conversion funnel? Where do customers abandon? etc.

b) Growth Equation — Next, determine your growth strategy and capture the metrics that matter the most for your growth. This can be summarized in a formula called “fundamental growth equation.” It will help you focus on the metrics that matter, as well as the features and initiatives when it comes to the idea phase.

• An example of a Growth Equation for a B2B site trying to drive leads for its sales team can be:

(website/landing pages traffic * email conversion rate * paid campaigns conversion rate * active user feeding one of the contact forms) + resurrected contacts = lead growth

c) North Star — Choose a North Star. This is the key metric of success (which can evolve with the product). The One Metric That Matters. The best is to have one (and only one) to narrow your focus and reduce risks. “The North Star should be the metric that most accurately captures the core value you create for your customers,” explains Sean Ellis… and doing so generates revenue growth.

d) Persona — Define your various target audiences through personas and understand their needs. To do so, start with the results of the data dive. It should give enough information for a first segmentation and to draw personas from it. Then, conduct interviews for each persona. You will need eight interviews for each persona.

2. IDEATE

At RevSquare, we use three tactics to ideate, which are inspired by the Design Thinking approach. Ideation needs to serve the North Star and the Growth Equation. Having ideas is great. Having ideas that matter for your growth is better.

a) Brainstorming — Some key elements for a good brainstorming:

• Involve your growth team. It’s a diverse one (more on growth team members to come).
• There are no more than 8 people.
• Make sure everybody does his homework BEFORE going to the brainstorming session, so make an assignment. Explain the goal of the session, the North Star, and the Growth Equation, and ask people to come up with a minimum of 10 ideas each. Volume counts in each brainstorming exercise. The more ideas, the more chances to come up with great ones.
• Make sure people do not criticize other ideas and be strict about this.

b) Interviews — The persona interviews is when you’ll gather user needs. Observe and try to figure out some latent needs. These are needs that are not expressed by the audience but that can be figured out by observation or analysis. Figuring out latent needs could give birth to major breakthroughs.

c) Surveys — There are two main types of surveys that you can use. A survey you link to or send via an email, or a survey within your application. Make sure the surveys are very brief. When it comes to pop-ups, make sure there are no more than two questions. Make sure they appear at the right moment (they can be very annoying when users are in the middle of something), and that you have created the right incentive. Open questions are very important. Instead of pushing your preconceived ideas, you want the user to tell you what he wants.

We also follow the idea document structure proposed in Hacking Growth, by Sean Ellis and Morgan Brown. We will share this document in an upcoming post.

3. PRIORITIZE

They are several ways to prioritize. We use different methods to do so. Here are four:

a) Refinement — Ideas need to be polished before scoring. After a brainstorming, the group does a fast first clean-up. Each member votes on each idea from 1 to 10 (lowest to highest). Then, take the average score and eliminate any ideas below 5.

b) Scoring — This is a simple gut-check measure for growth hackers. Use the ICE score concept proposed by the team of Hacking Growth. ICE = Impact. Confidence. Ease. Each criterion is also scored from 1 to 10, but the difference is that each idea is ONLY scored by the team member who suggested it. We’ll have more on the ICE score soon.

c) Idea Channel Fit — Define the optimal delivery channel for the idea. As Brian Balfour, the former Growth VP at Hubspot, says: “Products are built to fit with channels. Channels do not mold to products.” Channels define your product almost more than any other variable — products are tailored to perform within channels. So defining the channel cannot be an afterthought, but must be taken into account when refining ideas into testable increments.The channels being used for each of the ideas should be identified. Then, each of these channels is then scored.

d) User Stories — Once the ideas are ranked, user stories are produced to describe each of the ideas in the standard format. Growth ideas are small product changes characterized in a manner that’s consistent with the development team’s expectations. For growth ideas that are above and beyond a feature’s characteristics KPIs that measure the ideas are included as a type of acceptance criteria for success.

Note: Using scores to decide your priorities is great and convenient, but it can’t be the only criteria. For example, critical ideas for business growth are not that easy to implement. The Ease grade is really low, which gives the idea a low score. So, don’t only look at the score.

4. TEST

Now it’s time to feed your backlog and start testing. The four main components of the testing phase:

a) Sprint Planning — The Growth Sprint is part of the agile method in growth hacking. Growth Sprints follow a team’s capabilities and should ideally be run over a two-week cycle. A sprint backlog is based on team priorities that are decided with the help of various scoring techniques. KPIs are confirmed and cohorts/segmentations are prepared, if needed, for any email or social media campaigns.

b) Language Market Fit — Make sure that your application and your promotional material speak the same language. The language spoken by your audience. This applies to phrasing, terminology, and visuals.

c) Analytics Plumbing — Measure what matters. You know your KPIs, so make sure that you’re tracking the data that matters for each test you want to run.

d) Growth Meeting — At the end of each sprint, review the metrics, what was tested, and the key lessons learned. Select a growth test for the next sprint. Make sure you have enough ideas in the pipeline (if not, brainstorm).

These four steps — Analyze, Ideate, Prioritize, and Test — happen during each sprint. The ideation might be not necessary if you have enough ideas, but the prioritization needs to be informed by data collected during previous sprints. As in any agile, the first sprint is a sprint zero and more time is dedicated to the first 3 steps in order to produce a 3 to 6 months initial backlog.

The process outlined here was fully inspired by the fantastic book Hacking Growth. It is a must-read and a must-keep.