Tenon.io End of Year – Startup experience at 9 months in, product updates, and more
In September, I published Startup Experience: 6 months in. If you’re not already familiar with us, that might be a good prerequisite to getting context for some of the stuff below. This will be our “End of the Year” post, at 9 months of startup experience. I’m going to break this post into two sections: Business and Product.
Tenon continues to enjoy enthusiastic support by people in the accessibility field and our customers. Here’s what a few of our paying customers have to say:
Great functionality, superb customer service, sincere solicitation of customer input.University of Massachusetts Medical School
Accessibility is an important and often overlooked portion of web development. As a company who does a lot of work for the government accessibility is more than the right thing to do, it is a requirement.
Tenon makes it easy for us to automate a large portion of our accessibility testing. Between the website, gulp module, Visual Studio plugin and access to the API we can easily integrate testing throughout our build process.LMO Advertising
You can read more such quotes on our AngelList profile. This feedback speaks for itself in terms of validation of our product and service. We don’t take it for granted and we believe we can always do better, but we also truly feel that if our product helps our customers, we’re more than halfway there on attaining our goals.
In addition, organic traffic to the site is consistently up 13% month-over-month and inbound leads for Enterprise sales continue on a consistent basis.
I think every business owner says “we’re not doing as much business as we’d like”, so we’ll assume that’s a given.
In the last half of the year, we found ourselves burning money faster than we were earning it. Prior to this, we were actually making modest money but at least remained cash-positive. We made a few unwise decisions when it came to going “all in” on certain development tasks and wound up accumulating debt. In the grand scheme of things, the debt is relatively small and can be repaid over several months. The thing is that as a bootstrapped startup we’re particularly sensitive to debt so having it looming over our heads is pretty stressful.
We are experiencing 6% growth in user registrations month-over-month on zero advertising spend. This is lower-than-average growth. I’d like to see that number double, or even triple, but doing so will probably require spending money on ads. When we’ve experimented with advertising in the past, full conversion of users was too low to warrant the ad spend.
There are two major reasons for the low conversions: a weak landing page and unlimited free access to the demo on the landing page. The latter case is an interesting challenge. We know that there are people who use the demo on a regular basis as their way of using Tenon without paying for it. 11% of API usage is via the demo and most of that is from return visitors. The obvious downside is that we’re not getting paid for that usage and a pessimist could say that we’re being taken advantage of. On the other hand, we have converted some of those users. Both Fiserv and Kaplan were frequent Demo users who later converted to paid plans. We need to find the right balance of allowing people to demo the product without just giving it away.
To fully resolve these growth challenges and make advertising more successful we need to optimize our landing page quite a bit. As mentioned in the last update, we knew the landing page was a weakness. Asa made significant improvement to the landing page. Those changes have provided a demonstrable improvement in registrations, but we know it can be better. We want to aim for around double the amount of conversions on the landing page and that is a very attainable goal.
Lessons & Advice for other startups
As I mentioned in the previous update, after our launch we got too distracted by product development and expected sales to just appear out of thin air. This late attention to sales has had a lasting effect in that there are a lot of people we’re talking to now that we should’ve been talking to months ago. To other startups I say you need to do the opposite of what we’ve done in this area. Lock down your landing page now. Constantly A/B test the landing page and improve on it.
Another thing we focused on too late was our pricing strategy. Early on, we did some pretty informal research into what various products charge and then basically threw together what we felt were some fair prices. Pricing that way is stupid. We went into things not understanding what people felt was a good value for the product. So we decided to gather some data on the subject by putting up a pricing survey (you should take it!). Since the survey is new, we’re pretty low on data, but already learning new things.
If you’re bootstrapped like us, you have to watch expenses like a hawk. Unless you’re experiencing some sort of viral growth, managing expenses is just as important as getting new income. There are all sorts of awesome new tools and services out there that look useful for startups like CRMs and Help Desks and sales prospecting tools and marketing automation and the list goes on and on and when you add them together, they will eat your revenue like a bunch of parasites. For that reason, we’ve completely trimmed the fat on anything that doesn’t have a direct and clear improvement on revenue or product.
Did I mention that you should be selling? OK, just one more time: you don’t make money by creating an awesome product. You make money by selling your product. The product won’t sell itself. Trust me, I wish it would.
This section is so much easier to write. Product is fun. I want my tombstone to read “He loved his family and made cool shit for a living” because that’s what really excites me.
The product is stable and continues to grow in terms of quality and features. Since our official launch we’ve stayed above 99.7% uptime every month for a system that has supported several million API requests from over 2200 users from every country and every time zone on earth.
Tenon is unequivocally the only product that supports the entire web development lifecycle from linting the first code commit, to being added to the developer’s unit testing, QA automation, continuous integration, and content management. Take a look for yourself.
Together, Asa Baylus, Karl Groves and our small handful of developers have closed out 1341 user stories and issues and there’s nothing stopping our relentless pursuit of making the best web accessibility testing product out there. New features being pushed in the very near future are going to further broaden the list of reasons why Tenon is in a class by itself. It isn’t marketing speak when we talk about a “revolution”.
There are a handful of things we see as clear opportunities for improvement in our product:
Our development of new accessibility tests has been slowed as we concentrated on “big” features, some of which remain unfinished. Given our limited availability for new development having unfinished features drawing us in different directions means that we just have more and more stuff that needs finishing. Testing is one of them and is evidenced by our lack of color contrast testing. We regard color contrast testing to be something any accessibility testing tool should have and yet we don’t have ours finished. There are also quite a few additional tests we’d like to add and we need to get better at expanding these.
Our documentation of the features on the Tenon.io website is nonexistent and while our API documentation is complete and accurate, we’re not sure how user-friendly it is. Either way, both sets of documentation need to be improved. We suspect that one of the reasons why conversion is low is because people don’t understand all the things it can do.
Tenon is only capable of testing URLs that are publicly available. While you can send page source, some users want to be able to test their localhost or their development environment. Naturally if they purchased an Enterprise license they could place Tenon behind their firewall and there’s no problem. Unfortunately, the pricing and logistics for such a thing is out of some customers’ reach.
Tenon lacks a spider. On a personal level I think spidering – that is, automatically traversing through the site and testing each page – is like hitting a fly with a sledgehammer. I’m skeptical of the usefulness of the bulk collection of accessibility issues. That said, it is a request we frequently hear from users and potential customers.
Below are a few of the things we have coming just around the corner.
As I noted above, a major challenge for some users is the inability to test URLs in their development environment. With the new Tenon Tunnel, you can expose your local development environment to Tenon through a secure HTTPS connection. This is ideal for fully integrating Tenon in your unit testing and CI processes. For instance, if you’re familiar with using something like Karma/ Mocha/ Chai to do unit testing, you’ll be able to use Tenon in the same way, when combined with one of the Grunt tasks. You’d simply set up your Gruntfile.js to open the tunnel, run your tests, and close the connection. You’ll get fully DOM-aware testing of your local dev files. If using it as part of a Grunt task or unit testing isn’t your thing, you can use the CLI to open the connection which you can then use to get the Tunnel’s URL and test that using the Tenon website. Effectively, this completely eliminates the need for on-premises installs.
The Projects API will offer fully RESTful access to projects, including the ability to create spider projects. Spider projects can be run immediately or can be scheduled. Together with the Test API, you can use the Project API to set up a new Project for each build, test everything in that build, and report back using the Reports API (below)
The Reports API will be a read-only API that allows users to get over a dozen types of reports at the account, project, or page level, such as issue counts, top issues, tests passing vs. tests failing, and more. This can be used by customers to create custom dashboards or added into existing integrations.
This is one of those looming, unfinished features. This microservice can take a screenshot of the entire tested page as well as each individual issue location. The microservice itself has been done for a while, but not integrated into the main product.
Overall, this past year has provided us with excitement, anxiety, stress, and triumph. We’ve had a significant impact on our market and paved the way for future success. We look forward to learn from both our mistakes and our successes as we continue into 2016.