My perspective is that of a CEO of Wellzesta, Inc., a software company in the mobile health/wellness space. I function also as CTO, COO, and CFO. Our company produces an enterprise level App called Wellzesta Life that is offered through business-to-business (B2B) sale to providers in senior living.
In this article, I outline the stages that a typical software startup company goes through, give some advice, and relate a few personal stories. The article is intended for beginning entrepreneurs, experienced entrepreneurs (hopefully will get a few head nods), for persons that come in at the later stages of the company (and miss out on all the fun), and for the “just curious.”
Step 1. Initial concept
Have an original concept. For Wellzesta, it was measure wellness.
Develop the concept. Draw what the concept looks like. Pencil and paper wireframe drawings are fine.
Great , (0) you have an concept! The rest of this article is about (1) turning this initial idea into a product (2) that is sold to generate revenue (3) that is needed to have a sustainable business.
Reality check — Please re-read the last paragraph. If you are at ground (0) then I encourage you to read this article carefully. Commit to gaining the “practical training” in business that is required for you to succeed.
Enrolling in a technology incubator/accelerator is a great way to get this practical training. Wellzesta co-founder, Kyle Robinson, represented us at Groundwork Labs, a technology incubator in Durham, NC.
Tom Rau (developer intern) and Kyle Robinson (co-founder of Wellzesta) after winning the Groundwork Labs MVP award.
Incubators, accelerators, and the like are beholden to investors, which can include: the state or local office of economic development, benefactors, and angel investor groups. All investors are seeking a return on investment (ROI). Measures of success include: local jobs created, investment dollars secured, and the Big 2 financial scorecards — profit/loss statement and balance sheet. Accelerators provide seed capital in exchange for equity. Since accelerators have “skin in the game” I understand that the the experience is different than with incubators like Groundwork Labs that do not provide equity.
At the risk turning this into a human interest story I’d like to discuss the additional problem that Groundwork Labs was solving for Wellzesta. 5 years prior, the Spirit drove me into the wilderness of South Dakota to learn to trust God and to sharpen my spiritual disciplines, most notably patience — skills needed especially during Step 7 (below). While in South Dakota I developed a research program in single molecule biophysics that was funded by a $1.5M grant from the NIH. My employer, South Dakota State University (SDSU), refused to release this grant to another institution. This complicated our exit.
Kyle’s forward deployment in Durham, in 2015, served not only to pass important business know-how to me. It established a beachhead for us in the Southeast in terms of location, social network, and professional network. Having this beachhead (c.f., “you must make this fire very big,” Out of Africa, 1985) was a tremendous morale boost during our last days in South Dakota (c.f., U.S. military withdrawal from Saigon, April 1975). I will be forever grateful to the Dean and Provost for providing much needed “cover fire” during our withdrawal — the Dean being a particularly good shot.
Step 2. Discovery
Do your future customers think your idea is great? Talk to them. What aspects of your concept resonate with them? Kyle, given her industry relationships, did a great job with this.
Understand your future customer’s problem(s). It is better to offer a solution to a problem than a product.
Make sure there is market opportunity. What is the market size? Fill out the lean canvas 1-page business plan.
One of the most important factors in the success of the business is timing. Note: this has nothing to do with your idea. It has everything to do with market intelligence. Good businessmen recognize market opportunity. Bad businessmen ignore the time element of market opportunity.
From Gartner (2013)
Look over the fence. Do your homework on your competitors. Your investors will. Why? Because it is important. You are proposing to occupy a space in the marketplace. Is this new space that is unknown to your competitors (a blue ocean), or existing space that is being poorly served?
Define the specifications for the minimum viable product (MVP).
Step 3. Buy-in
Have business model; need money. Raise money through any of the following:
1. Pitch the business model to investors —the so-callled friends, family, and fools (FFF) round.
2. Use personal savings.
3. [Good] Win a great idea competition (e.g., NC-IDEA). Awards are ~$50,000.
4. [Better] Secure a grant. Our SBIR grant to the NIH was not successful. Unless you already have a prototype that is in beta (Step 5) then success is unlikely. My advice: do not submit at this stage; doing so takes precious time that will slow you down.
5. [Best] Find a few customers who are willing to pre-pay — pre-sales. This is faster than getting a grant, allows founders to preserve equity, and establishes product-market fit.
6. Be prepared for a lot of NO’s. Don’t take the NO’s personally; this is business. The founder of Pandora suffered no less than 300 rejections from venture capitalists (VC’s) before securing funding. Tenacity helps.
Step 4. Production
Before developing the software, give careful attention to team composition and technology stack. Your choices here will strongly influence your likelihood of success.
Step 4a. Technology stack
Choose your technology stack carefully. You will have to live with it for a long time. For Wellzesta, Step 4a began during Step 1 since recent technical advancements made our concept feasible. You want a stack that can take you to stable release (Step 7) as quickly as possible. Also, pay attention to how your technology stack will scale, in terms of cost and performance, during Step 9. You must weigh new-and-cool against battle-tested. The engineering mindset leans towards the latter. For cloud computing your choices are:
1. None. Use on premises data center. Gasp!
2. Platform as a Service (PaaS, aka bare metal as a service). Examples: Digital ocean, AWS, Heroku.
3. Infrastructure as a Service (IaaS). Examples: Google Cloud, Microsoft Azure.
4. Back End as a Service (BaaS). Examples: Firebase, Kinvey, Realm.io.
We opted for BaaS with all logic on the client using the ember.js MVC framework. The runner up architecture was Ruby on Rails on either PaaS or IaaS.
Wellzesta Life is a mobile App. In mobile, the choice is native or web. We opted for native. The decision wasn’t easy. You want to be device agnostic, but you also have to perform. In the end, performance won out. We needed a racehorse, and we went with the best: iOS.
Step 4b. Choose the right team
The technology stack that you have chosen will determine how you write the job requirements. We needed an ember developer, not an angular developer. We needed an iOS developer, not an Android developer. Etc.
But team composition is much more than technical know how. The team members need to have excellent communication skills. They need to get along with each other and with you. Also, you need to get buy-in on the communications and productivity tools.
Culture matters. You want a team that gets the nuances of language. Time zone is important. If your developers are in south east Asia, you face cultural differences and a significant time offset.
Cost is a consideration. Cost for designers/developers ranges from $300/hr to $50/hr.
Team size is a consideration. You want the software built as soon as possible. Having more hands should help. But, be careful. When you increase team size, you increase communication overhead [thanks to James Branigan at Bright Wolf for pointing this out during the distillation of 5 years of experience into a 1 hr lecture].
Working with a dev shop offers a lot of advantages. You get a team with a track record of product development, an established culture, and a standardized productivity/communication toolset. You get flexibility — team members can cycle on and off the team during the development and maintenance stages of the software lifecycle. The more experienced developers will give you push back when your expectations are off base. If you are lucky, the dev shop is also an agency that is interested in business value and process. HE:labs in Brazil presented a strong value proposition in terms of cost, flexibility, culture, time zone, productivity, and business acumen. Wellzesta is glad to have partnered with HE:labs to develop Wellzesta Life. Don’t overlook South America as a source for high tech talent.
If you don’t feel confident about choosing a team yourself, you can work with a company like Designli who will find a team for you. Hold on: Are you sure you want to outsource such an important decision? You are a software company, right?
Step 4c. Develop the software
The entrepreneur should be as hands-on as possible during software development. You are in project management mode. In Step 2, you made sure that you were in the right forest. Here, your role as project manager is to make sure that saw is sharp. You should be up a few hours before the rest of the team, making decisions about the day’s priorities.
Look under the hood once in a while. The project manager should examine the code base to ensure solid construction. Remember:
You inherit all technical debt.
Our productivity tools consist of: Sketch, Zeplin, InVision, Slack, Trello, Basecamp, Google Hangouts, and Github.
Suggestions:
1. Use agile methodology. Avoid waterfall.
2. Involve as many members of the team as possible in drawing sessions that define “user flows.” The project manager must provide scope for these sessions. Project manager defines what the user “wants to do” (function); the team draws the user experience (UX), or flow, to accomplish the function. The capable project manager knows the business value of each function, so can prioritize development to create the most business value.
3. Pay attention to team morale. Our weekly 30 minute retrospective sessions were a lubricant and a tonic. Software development is intense. Some things go well. Some things don’t go well. Celebrate victories. Often, just saying out loud that something aggravated you allows you to mend emotionally.
4. Do usability testing early and often. Test your assumptions quickly.
5. Bad ideas are fine; just make sure that your bad ideas “fail fast.”
Speed and agility — the currency of a startup
Question: Should you spend 8 months writing, submitting, and waiting on the results of a SBIR phase I grant (Step 3) that will provide $50,000 — about 1/4 of your software development costs (Step 4)? Answer: No. Resist anything that will delay your progress to Step 8.
Question: Should you have advisors who are willing to meet with you once a month? Answer: No. The competitive advantages of a startup company are speed, agility, and adrenaline. Startups do things “fast and furious.” Decisions are made in seconds-minutes, not weeks-months. Kindly decline such offers. In our case, I was relieved when this bad idea failed fast.
Review
Before heading to Step 5, let’s review two important points.
1. Market research (Step 2) comes before financing (Step 3).
2. Financing (Step 3) comes before software development (Step 4).
Step 5. Initial deployment
Nothing beats real field testing. You wanted customer feedback? You got customer feedback.
Make your early adopters happy.
Begin focusing on marketing.
You will need a support portal/knowledge base/help desk. We tried out desk, zendesk, and intercom. We went with Zendesk. Zendesk comes with plans that included a knowledge base. We rolled our own using the KnowAll Theme for WordPress. This is a mobile first theme that is tuned for creating a knowledge base. We use WPEngine to host our WordPress sites, and we are very satisfied with them. All traffic passes through CloudFlare, which offers security, hot swapping (if needed) and a blazing fast distributed DNS. We use Google Analytics for site analytics.
Grow your professional network and your sales network. You will need them in steps 8 and 9. Become a power user of LinkedIn.
Step 6. Patch up the initial deployment
Congratulations, you are now doing updates on a live system. Friendly reminder: do daily backups.
Until now, all feedback was friendly. Bug reports and usability questions can be rather emotional and have a tone of urgency. If you didn’t set up the help desk and support ticket system (Step 5) then God help you. Because of your forward thinking, your support staff can remind the user that the answer is on page 5 of document x on the help desk.
A whole new set of emoji appeared on our Slack channel— :shipit: :thinking_face: :scared: :sweat: :crossed_finger:
You now have a mixture of new feature requests and bugs. Prioritize development in this order:
1. Crash fixes
2. Bug fixes
3. Tweaks
4. New features
Avoid the “new feature” death trap
Prioritize feature requests. Give preference to features that will scale. Exercising good judgement means saying NO to certain requests. The pitfall at this stage is being pulled out of the Magic Quadrant (top right, LEADER) by loosing your focus (CHALLENGER) or failing to execute (VISIONARY).
From Gartner (2013)
Collect user testimonials. Ask your early adopters for referrals.
Step 7. Stable release
Pay down technical debt. Fix those hard to find bugs. Refactor code. Write DRY code if you didn’t before. Issue the stable release when you are “unable to break the app.”
From the #wellzesta-staff channel on Slack.
The stable release is a major milestone. This phase requires strong leadership and patience that, if you are anointed, God has either given you or ground into you. Refactoring code means breaking code. The house will be a mess.
Non-technical members of your team and your customers may see little progress and become frustrated. The developers, the marketing/sales team, and your customers may pressure you to publish before ready. Steady the course! Expect your life to be rather lonely during this stage. You need grit and steadfastness in ample supply.
Imitate those who through faith and patience inherit what has been promised. (Hebrews 6:12)
The stable release demonstrates the ability to execute. Having vision is one thing. Implementing the vision is a separate matter.
The number one gotcha in our code was failure to check “if x != NULL.”
[Update] Step 7b. Patch up the stable deployment
In the original design we unknowingly made three assumptions that turned out to be false:
1. Users would be willing to type in a 4 digit PIN to protect their wellness data
2. Users could close out of an iOS app if asked to do so
3. Net user experience would be a linear weighting of positive and negative experiences
Missing the mark on these three points was a very painful realization. When your expectations are not aligned with reality, you have a choice: change your expectations or change reality. I suggest you pursue the former option. Swallow hard. Accept reality — don’t judge; don’t preach. Yes, eliminating the 4 digit PIN poses a security risk. Save the discussion for later. Your goal is to eliminate friction, not to be right.
Step 8. Early Sales
Congratulations, you have a stable release. You are now in sales mode. You are now in a sprint to gain a strong foothold in your primary market. This means having 5 customers. Sales not only brings in revenue, but helps with marketing.
Steps 7 and 8 should be done simultaneously, if possible. Collectively, they de-risk the business model. Steps 7 and 8 demand different sides of the brain. I am unable to multitask in this way.
Transitioning from tech geek to product evangelist requires serious mental retooling.
Probably the most helpful thing a CEO can do at this stage is generate PR. PR can come from invited talks, visiting potential business partners, blog posts (like this one), publishing results (science!), giving interviews.
Get in front of people!
Get the word out about your great product.
Step 9. Growth
Congratulations, you have exited the valley of death. Time to scale.
Development. Add new features. UI tweaks. Develop plug-ins
Financing. Secure Series A funding from venture capitalists. Investors like to inject capital when the business model has been sufficiently de-risked (proven) and the injection of capital can lead to large returns (sales).
Sales. The factory has been built. Now it’s all about sales.
Step 10. Exit
Choices are:
1. If you are not too exhausted from Steps 1–9 then don’t exit: become a lifestyle business and coast a bit. If you have investors, they will not like this option.
2. Scale within your market or into new markets. Talk to the growth equity folks. Need a minimum of $1.5M/yr in revenue. Hopefully founders still own most of the stock. By taking on growth equity, you are selling your shares = converting stock into cash. Hope you have a good valuation.
3. Be acquired. If this is your destiny then you should have been talking with buyers in Step 9 (or hired someone to talk to buyers).
4. Go public. It is possible.
John Robinson
Health & Wellness. Big Data.