Category Archives: marketing

Marketing for Engineers (aka Growth Hacking)

So you are building a new product. Great! However, one of the worst mistakes you can make is to wait for it to be done before thinking about how you are going to market it. You should be thinking about how you will reach users and convert them into customers at every step of your development process.

The following is a basic framework you can use to think about marketing your product that you can consider at every step along the way.

1. Practice Telling Your Story

The first thing you should do, before even starting to build your product, is to understand how you will explain it to your customers. What problem does it solve? How does it work? Why is it the best option? Amazon takes this even farther, asking product managers to write the press release for the product before they start building it. Working backwards forces you to think about the product from the perspective of customers, instead of building a product just for the sake of building.

You should test this story by telling it to potential customers continuously during development. When you have wireframes, show them. When you have a prototype, demonstrate it. Do not wait until the product is “done”, because by then it might be too late to learn how it needs to change to match the story that customers want to hear.

You will know that you have mastered the storytelling for your product when describing it to potential customers comes naturally and the product tells most of the story for you.

2. Model Customer Conversions

It is easy to fall into the trap of optimizing your user interface for expert users of your product. You are an expert user of your own product, after all, and it is natural to want to optimize it for yourself. However, the most important feature of any product is the first time user experience  (which is how the customer becomes a customer in the first place). If you don’t obsess over the initial experience, customers may give up and never make it to the the point where your expert user interface has a chance to shine.

You can think about your first time experience as a sequence of events. For example:


This example sequence is simple, with the blue circles representing the actions we want the user to take and the gray circles representing alternatives. For this product, the ideal first time experience is as follows: the user clicks on a link, signups up for the service, stays active for one week (is retained) and eventually makes a purchase. At each step the user could have failed to complete your preferred action (instead of signing up for the service they abandon the registration page) and this represents a lost opportunity. By understanding this flow you can understand how you expect customers to progress from prospects to ideal customers and where you might lose them along the way.

Of course the first time user experience for your application will be different, but it is composed of some number of steps that you can set up in the same way. There are some great examples of first time user experiences on FirstTimeUX. Your goal is to figure out these steps and make sure you can measure them effectively using your analytics solution.

The very last step should be the ideal behavior of an ideal customer (full purchase, invite 10 friends, etc.) such that a user that completes all the steps in your flow is your ideal customer. By measuring how many users achieve each step you will understand how effectively you convert prospects (clicks) into ideal customers. If you are losing customers along the way you should know quickly and be able to plug that hole.

3. Plan Your Marketing Channels

When you were mapping out the customer conversion flow in Step 2 you likely thought a lot about where the customer might be coming from. Where did they see the link that lead to that click? You have many options for getting users to that first step in the conversion flow including:

  • Content marketing (blogging, tweeting, press)
  • Social Sharing (Facebook, Pinterest)
  • Invites from friends
  • Paid advertising

The conversion flow for a customer coming from all of your channels may be the same, or they may all be different. You should model conversions based on the experience you think is most likely to convert a prospect from a given channel into an ideal customer.

Many people will dismiss paid advertising and believe that your product is so great that it will succeed through word of mouth and virality alone. Alas that is rarely the case. Even Google spent $12B in traffic acquisition costs in 2013, and their product is nearly ubiquitous. It is important that you do not overspend using paid advertising (see The Most Important Equation for Your Business) but it almost always a part of a successful marketing campaign.

Be as creative as possible in identifying possible marketing channels and how your conversion flow might differ based on the source of the click. Many of the channels you identify will not be effective and you will discard them, so you want to cast a wide net.

4. Measure, Rinse, Repeat

As early as possible, you should begin testing your story, conversion model and marketing channels. It is almost impossible to be sure which channels will be the most cost effective, how users will convert or what story will resonate the most. Do not wait for your launch to start figuring out how best to attract customers. Some of the best companies will test their marketing channels during their Alpha or Beta tests, while others will “soft launch” their product in other countries to test their assumptions.

When you do start testing, be sure to track performance very closely so you can make decisions later. One way to do this is by compiling your marketing channel performance in a spreadsheet. An example template for such tracking is below.

Marketing Channel Tracking Template

This particular template ties channel performance directly back to cost. This may seem strange, especially for non-paid promotional channels, but it is important to remember that nothing is free. Writing blog posts and soliciting the press for articles seems free, but takes up your time which can be expensive. Building social channels within your product can lead to word of mouth but takes away from other features. In the end, everything you do costs money and you need to understand the return you get for that investment.

When you are running at full stream, you will not track your marketing through a spreadsheet and instead will likely use one of the many advertising tracking services available. However, in the early days it can be helpful to do it yourself by hand so that you really understand what works and why.

Conclusion: Hire a Marketer

While this is a great start, none of this is a replacement for hiring an experienced marketer. Software engineers tend to discount the value of marketing because it is so different from the hard skills of computer science. However, a great marketer will help you explain to your customers how amazing the product is that you have built. Marketing is the act of understanding your customers and communicating with them in a way they understand. Without strong marketing, a great product may die in obscurity.

During your product development, always consider how the product and marketing strategy work together. In the end, they really are the same thing.

You Need 10 Times More Customers

As an engineer, one of the first things you are taught is how to break down a problem into smaller and more manageable problems. Want to build a bridge? First, you’ll need to design the supports and the connections with existing roads, then the external structure. Building a house? You start with the foundation, then the frame, then the plumbing.

Building a company is no different. Billion dollar companies are not built overnight, they are built through a progressive series of steps. If you attempted to envision a billion dollar company from day zero, you would be overwhelmed by the challenge.

Take the following example of a high growth company’s life cycle:

Growth Stage Company

This is an intimidating chart. The number of customers grows slowly at first and then hits an inflection point where it begins to grow exponentially. How in the world do you design a company that can grow this way? How can you plan for that kind of growth?

The short answer is that you can’t. You can get lucky and have it just happen, but that is so unlikely that you would be better off playing the lottery.

What you can do, is break down the problem into manageable phases:

Growth Stage Company - Milestones

Here the growth path is broken into four stages: Series A, B, C and D. These might correspond to your funding rounds or they might just be milestones that you set for your company. At each stage, you have acquired 10 times the customers that you had at the previous stage. While these are still aggressive goals, they are goals that you can achieve.

The key lesson here is not that you should plan out the entire lifetime of your company on day one. The lesson is that, whatever stage you might be at now, your next goal should be 10 times the customers you have today. Setting such a goal will force you to take a hard look at your company and think about what will need to change to achieve that next level. When passing from one level the next, the following are typically true:

  • You will need to start doing things you have not done before.
  • You will need to stop doing some things that helped get you this far, as they will start to become liabilities.
  • You will need to continue doing what makes your company unique, the core of your identity.

What those things are will depend on the business and the level you are at today, but asking the questions will help you position the company for success. Even if you fail to achieve 10 times customer growth, you will have positioned the company well for growth and increased the overall value of your company.

Why should you aspire to being a high growth company and plan this way? It is true that slow growth companies can be very successful and often feel much more comfortable. However, the vast majority of wealth created in entrepreneurship comes from the high growth stages of business. It is during that high growth stage that your capital efficiency will be at its highest and hence the value of your company appreciate the most for the shareholders. For publicly traded companies, this is why high growth companies command a premium on their share price.

So, even if you are just starting your company today, think about how you will reach the next level of growth. Before you know it, you might have achieve that milestone and be thinking about how you grow by another 10 times.

How to Avoid Naming Your Company “Brain Rocket”

Naming something is very, very hard. Not only is it hard to choose a distinctive name that describes your product or company, almost all of the good names are already taken. There are naming firms who charge hundreds of thousands of dollars to help large companies choose new names, and even then they often end in disaster. So, how should you pick a name?

While there is no silver bullet, it helps to have a process. In addition to reviewing the process, we will practice on a fictitious company and try to give it a name. For this exercise, let us assume we are starting a new consulting company that will help set up IT infrastructure for small businesses. 

Step 1. Metaphor
All of the best names are based on metaphors that describe what the product or company does. Literal names (such as The IT Company) are rarely distinctive enough to be remembered and easily confused with other, similar companies. The more creative your metaphors the more distinctive your name and the easier the naming process will be in the future.

For our example IT consulting company, what metaphors work? Since IT touches almost all aspects of business these days, the company will provide a wide range of services to help smaller businesses grow. Any metaphor that involves getting bigger over time would work well here, including construction and stacking. The word “grow” is interesting and seems to fit well with gardening or farming so let us choose the metaphor of gardening.

Step 2. Word Association

It’s time to exercise your creative muscles and play word association with the metaphors you chose in Step 1. The goal is to create a large list of words that relate to your metaphor, as many as you can. If your creative muscles are out of shape you can use online tools to help generate a list of source words related to your metaphor.

Based on our gardening metaphor you might come up with the following list:

  • Seeds
  • Sprouting
  • Green
  • Watering
  • Harvest

Some companies choose to include more descriptive words in their name to avoid customer confusion. It also makes finding a domain name significantly easier since all single words are taken. Some words that describe IT infrastructure:

  • Technology
  • Infrastructure
  • Hardware
  • Networking
  • Installation

Step 3. Try some names

It’s time to combine the words from our word association exercise, creating a list of potential names. It is okay if they are horrible since we’ll filter them out later. Note that combining the words into new names can take any form you like and sometimes the best names come from combining two words into one. For this exercise we won’t invent new words, just create compound names:

  • Seed Technology
  • Harvest Installations
  • Sprout Networking
  • Green Hardware

Of those, “Green Hardware” and “Seed Technology” sound like biotech companies so let’s throw that out. That leaves us with two potential names:

  • Sprout Networking
  • Harvest Installations

At this point you need to be practical. Check available domain names to make sure they are available and/or affordable. Use a service like 99 Designs to have some example logos created for your names to see how it looks. Talk to your friends and associates about the names and get some input.

It is possible that after all of this you end up with no good names, forcing you to start over again. For a given product I typically find I have to repeat this process 2-3 times before I end up with some really good candidates. If you are lucky you will find a name you really love, but often you will find a name you can learn to love.

I hope that helps. You could always just fall back on naming it after yourself (it worked for Bloomberg) and there isn’t really anything wrong with that. In the end if you have a great product or service, treat your customers well and deliver on your promises it won’t matter what your name is – people will remember it.

NOTE: This post is adapted from an answer I posted on Quora in Startup Advice and Strategy.