What went right, wrong and was missing from our service business

July 8, 2017

After over two years working in the public sector, I’ve had the opportunity to look back and consider what went right, what went wrong and what was missing in our previous company.

In government, I’ve grown substantially as a designer. Daily exposure to the internal mechanisms of the organization and knowing now what it takes to ship and support digital services has challenged me to greatly broaden my definition of “design”.

It’s been working in this environment that I’ve developed many new opinions and ideals and been able reflect on some of our previous choices. If I were to start a service business all over again, there’s some key things I’d do differently.

1. Make a better offer

For the eight years Subvert was in business, we focused solely on digital design and development services. These services included making websites, mobile apps and transactional web applications, all sold to clients on a per-project basis.

We did good work, but our offer was far from unique. There’s many, many other companies, big and small, local and elsewhere who do the same thing. And that’s okay. Projects are generally plentiful and we created a relatively stable business with this model. Problem is, we were really just scratching the surface compared to what we could be doing, scope of service-wise.

To paint a picture, let’s consider one of the larger American digital agencies and their scope of services. They have over 70 employees and this is their stated list of capabilities:

“KeyKey capabilities of a leading digital agency

Almost all of these capabilities relate to a per-project based business model. The work is centered around making the ship, not making, guiding and then keeping that ship afloat during its lifetime of adventures through vast, turbulent waters.

Most of our clients were repeat customers, but looking back now, I don’t think we went even close to deep enough with them. We didn’t offer the truly high value services they really need for us to be considered long-term essential partners in their success. There’s certainly some companies that do, but we weren’t one of them.

After having been in what’s occasionally a client role at government, I now perceive these truly high value services – what I would call the “hard work” services - to be:

Organizational policy and process design

This is the research, advice and assistance that clients need to analyze and then improve their internal processes through changes to or creation of new organizational policies, including governance.

Our company was never hired to improve internal policies or create substantial new processes, including the elimination of poorly performing versions. We simply created digital products (websites and applications) that interpreted existing policies and made improvements at an often very small and removed level of impact.

Policies, governance and business processes are what really create a big impact within a large organization, not digital products. Products are simply a related outcome.

User research, observation and product implementation

This is the organization and facilitation of user research and observation sessions, followed up with an executive report and a detailed plan for making changes.

Subvert was occasionally hired to conduct user research, observe and implement, but not enough of it or in a way that made the best use of the results that we produced or set our clients up for success.

Having now been on the inside, I see that actually getting a potential improvement noted, documented and then implemented into an existing system takes a significant amount of focus and effort. You can’t just leave the baby basket on the doorstep, walk away and expect they’ll be okay with their new parents.

No, instead these issues should be added and prioritized in an internal product backlog, not just in the report. Issues always have to easily understood and actionable.

As a consultant, it’s often very difficult to make suggestions and then be able to directly improve the product itself, especially if it’s behind walls of policies, permissions and piles of technology. Being able to drill through all of those layers takes the right combination of skills, relationships and internal leadership decisions, so you need to equip the client in this way, if they are not already.

Prototyping and integrated systems development

I've mentioned before that in my current role at the public service the act of designing and programming a digital service often represents less than 10% of the total work it involves to get that new service out the door and into citizen’s hands.

“WhatWhat lies beneath

The remaining 90% is navigating the service through process and policy as it matures, evolves and gets its boxes ticked.

That said, the work of building and iterating on a functional prototype, then integrating and supporting a more final version with hooks into existing enterprise systems (usually via web services) is still an essential part of the program.

At Subvert, we sometimes built applications that stood alone in their own environment, but that’s not what internal IT administrators and system end-users appreciate in the long-term.

Despite their potential positives (better usability and a more beautiful interface, for instance) purely standalone applications make all of their work harder and more complicated, as you’re creating yet another application that needs to be maintained, managed and paid for over the long-term. I’ve found that the more disconnected something is, the more costs it produces.

It’s much wiser to make friends and work in a cooperative, integrated manner than create enemies who sigh when they hear your name in the hallways.

2. Build a complementary, open source product to sell and support

After Mike joined Subvert in October 2009, we immediately started planning the development of products we could sell and support. Our first venture was a community-focused website that sought (and failed) to make money through advertising. After that came several Windows desktop and Windows Phone applications, then an iOS-based product.

None of these products did particularly well.

In 2011, we started working on what would become Ibbit. Three years later in 2014, we had created a solid and desirable system. We hadn’t yet achieved what I’d consider product/market fit, but we weren’t that far off. Unfortunately, the business was suffering in other ways and we simply ran out of financial runway.

We today continue to support our clients who still run their online businesses with Ibbit. Features are just simply more slowly added and changes aren’t as quickly made as they once were, but the product is still very much alive and breathing. If I were able to go back in time to perhaps 2013, I’d do a few things differently with Ibbit.

First, I’d make it open source by default so that others could improve it and we’d gain more exposure for the product.

Secondly, I’d sell paid support and consulting services for Ibbit, rather than rely on a commercial subscription-based licensing model like we’d implemented. Those subscriptions were working in terms of revenue generation, but I think we could have done better and been smarter with the offer.

3. Run the business as cheaply as possible

After Subvert got off the ground and was making money on a regular basis, I started paying myself more. After all, I had a family to support and ends to make meet. When Mike joined me in the business, we continued to pay ourselves reasonably well; him for the same reasons.

Looking back, I wouldn’t change the size of our paycheques but I’d certainly scale back on most of our other expenses. The trimming and cutbacks we started making a year before Subvert shut down were adjustments we should have made much, much earlier in its operation. Those savings were and should have been to:

Rent a cheaper office

For example, the office we moved into during 2010 cost us, on average $2,575 per month. That’s $30,900 per year. Comparatively, the first office we were in was just over $900 a month, or $10,800 per year. We probably should have stayed in the first office for a much longer period.

There were also zero co-working spaces in Whitehorse back in 2010, but now there are several and often they are full of diversely skilled people. A dedicated desk at (co)space costs $279 per month, which for us would have $558 per month or $6,696 per year.

We sure could have used that extra $24,000 annually when times were at their leanest.

Go cheaper, on everything else

Cheaper insurance. Less monthly subscriptions. Fewer contractors. Plainer furniture. Older equipment. No leases (own everything in case you need to suddenly liquify). More economical professional fees. Lower grade coffee.

No, not that last one. I’m not willing to skimp on lousy coffee. Life’s too short for bad beans.

There would have been many ways for us to save more money, but it was too late. We made cuts all over the place, but we should have started making them sooner.

Put away a fixed amount every month, and then don’t touch it

As the years went by, I moved on from the model of regularly squirreling money away into a high interest savings account, but I really should have kept that practice going. Instead of reaching into that account when we hit rough patches and slowly chip away at what we had saved until there wasn’t much left, I should have let it sit.

Instead of withdrawing small amounts here and there, it would have been much wiser to simply tighten our belts, cut other costs and hold onto that money until when we really needed it. The trouble becomes it can too often feel like you really need it in that moment, when in truth you don’t and you just need to ride out the lull.

4. Always be marketing

Subvert got most of its work through previous client relationships, word-of-mouth and our website, especially the blog.

Keeping our sales pipeline more consistently full is something at which we could have worked harder. I would go through periods where I’d concentrate more on marketing, but those efforts weren’t as consistent and strategically managed as they had to be.

Looking back, I should have done more regular, highly-focused marketing and promotion of both Subvert and our product, Ibbit. In hindsight, better search engine landing pages, more email newsletters and a bigger presence on social media would have done wonders for driving traffic to and interest in what Subvert had to offer.

In the end

I don’t regret anything about starting and running a company for the eight years that I did. I was afforded opportunities and experiences I would have never had if I didn’t take that initial leap of faith and then work hard to keep it going. I’m very proud of and look back fondly on what we achieved in the time that we had.

If you’re thinking of starting or are currently running a service-based business, hopefully these insights can help you find your own path to a more efficient, profitable and personally fulfilling operation.

Contact me if you have questions, comments or want to share your experience.