A year ago, I wrote a tiny article about marketing tools that helped us win the first 5 customers. The stack and the strategy were sort of successful because 12 months later we crossed 100. But we did apply some essential changes too and so, I’d like to share a more thorough summary covering both the tools and the tactics in this article.
To give you some perspective about our business: Voucherify is a developer-friendly, API first promotion platform. It offers programmable building blocks for coupon, referral, and loyalty campaigns.
From the very beginning, we set out to be focused on developers and, therefore, on BS-free marketing. Our role models included popular API-first platforms we’ve used before and liked working with on previous projects. By analyzing Stripe, Contentful, or Shippo open documentation, we learned how they design the API, onboard users, or handle support cases. All this helped us to shape the first product vision. But what wasn’t so clear was what marketing steps we should take.
The role models are usually named as such that way because they are successful. And when you first hear about a successful company in the tech world, their original marketing tactics are usually long gone. To make matters worse, it’s not always easy to follow the history of their marketing strategy in the early days. Finally, we don’t build the same damn thing, so our go-to-market strategy has to differ in some ways. All in all, as there were no clear-cut marketing tips, we had no choice but to start experimenting. Firstly - what worked.
This one turned out to be our power lead generator for Voucherify. Climbing up in SERPs (search engine result page) got us more and more sign-ups on a daily basis. We managed to put ourselves on the mythical first page for our most valuable keywords such as coupon software, voucher software, coupon API, etc. And since then, there hasn’t been a single day that we didn’t have more than 10 new users on the platform. I can’t stress enough how important SEO is for us at this stage.
SEO equals content
It’s common knowledge that the more frequently your content is shared, the better rankings you get. And our results confirm this. Although it’s hard to find the precise correlation between the particular types of content and the number of backlinks, we can see that a continuous create/promote cycle yields better and better outcome. It reminds us of an avalanche effect.
“But what content should we produce then?” - We were asking ourselves. It wasn’t clear in the beginning, and it’s still not totally clear now. What I can say is that we split content into 2 categories:
For marketers - blog posts which showcase our platform, promo campaign best practices, white papers, product updates, and guest posts.
- help us get better ranking in SERP
- can be reused for onboarding and support
- Some of them get outdated as we iterate on the product
- Some more technical posts, e.g. about automating coupon send out with email service providers are just too technical for marketers to find the value immediately, ergo they’re not shared. Saying this, I’m wondering how the heck the AI, VR, IoT gang (way more complex stuff than our simple promotion API) get shared so often.
For developers - we gained our tech experience when working with e-commerce startups. No particular domain, general high-to-mid level articles that will hopefully get the attention of architects/CTOs.
- We’re devs ourselves, so it’s easier for us to write and promote tech articles
- 20 senior devs have a helluva lof of experience and therefore a lot of material for new posts
- Shares ratio looks fine
- It’s hard to “convince” a senior dev to write a blog post
- Less helpful when it comes to SEO (keyword optimization)
- High-level articles are not as happily shared as brand new React features
In general, what we want to publish in the long run is popular content for marketers who are not afraid of technology and developers who understand business/economy. It’s a little hard to do this because there are far fewer people with these skills than say, regular marketers or developers. Moreover, these people don’t “share” so often. Anyway, we believe we’ll finally get there by iterating long enough and accumulating more and more interesting insights along the way.
Not only a blog
We also ran a small experiment with an open source contribution; We released a subset of our platform which generates random codes as a free library with a promo link to our main platform. Some developers wanting to issue random coupon codes might be interested in redemption, tracking, and the whole promotion management process as well, right?
In the second iteration, we decided to make this library available for marketers too. And this is how the free coupon generator was born. It turned out to be a nice lead source as it quickly ended up on the first page of Google.
Tools we use:
- Unamo (formerly Positionly) - keyword and backlinks monitoring
- A free trial of Moz or DeepCrawl to find SEO bugs
- Omnigraffle and Piktocharts for diagrams and infographics
- Traffic analysis: Google Analytics, Heapanalytics, hotjar
SEO is really only a way to draw early attention before the next, more essential marketing process kicks in - product marketing. This includes all artifacts that help our customers understand the value behind our product. The range of product marketing activities is wide, but we split them into 2 categories for convenience:
- Website and landing pages - still not quite ready to invest in a full-blown, beautiful, custom website like Stripe, but with Webflow you can still build quite a nice and interactive one.
- Instructional blog posts - example, we host our blog on Webflow too, however note that we mix visual and text messaging to satisfy both users and Google bots.
- Help Center - a marketer-oriented knowledge base, a lot of self-help tutorials which promote other features at the same time. Apart from a ticketing system, HelpScout provides a very pragmatic and agile knowledge base platform. Key features: price, usage/search analytics, rich-text editor, API, “embed in the app” option
- Open docs & examples - as an API-first provider, we initially need to convince developers. Therefore we gradually added more and more tutorials, how to’s, and examples to the specification. And the examples are more and more illustrative and interactive. For example, we use Postman to demonstrate what our API is capable of doing quickly. Similarly, we tapped into the Glitch platform which allows you to create front- and back-end applications with the web-based editor only. The most interesting part is that you can quickly spin off a copy (remix) of the example and tweak it yourself. So, a developer can try out our API in minutes, no sign-ups or boilerplate needed.
- Content newsletters - we maintain 2 newsletters: one for the general marketing posts and the second one for our “How to build an online marketplace” series.
- Newsletter - it’s a pretty standard one, I guess. We share product updates and interesting reads we published on the blog. As with the blog content, we have two target groups (devs and marketers) interested in different topics and so, we might try issuing 2 separate newsletters in future.
- Interactive onboarding tutorial - this is again a standard drip campaign. Nevertheless, we get the feeling we should approach this topic in a more data-driven way.
For both tasks we use Intercom. It’s a super flexible tool with an API and a usable mobile app but the quality drops are noticeable often. Look at their incident history; there’s an issue almost every couple of days. On top of that, they are pricey and not early-stage friendly. Admittedly, they’ve introduced a discount program recently but you have to be a new customer. (We’re probably going to replace it - if you know of an alternative, let me know in comments.)
Let’s sum up the stack:
- Webflow replaced Squarespace
- We moved non-dev documentation from Readme to HelpScout
- Glitch and Postman enhanced the developer documentation with interactive, easy-to-use and easy-to-hack examples
- Intercom and MailChimp handle newsletters
Product is the king
Finally, the SEO and Product Marketing constitute a runway for the third and most important marketing task - product development. This is where we put the greatest amount of energy and I believe it will stay like this forever. Perhaps, when we discover a new, guaranteed sales channel, our marketing effort funnel sections will change their diameters but the shape will stay:
The fantastic thing about product development in the early stage is that what you really need is a set of basic cheap tools like whiteboard to come up with new features and Email/Skype to verify these ideas with your customers. That’s exactly what Tom (our CPO) does alternately conceiving features with the dev team and spending countless hours on the phone/skype.
Of course, it’s also worth wrapping your head around quantitative data. This is why we’ve recently set up Metabase (an open source data analysis tool) but you can also get the same results by simply querying the database (assuming you know SQL). Anyway, the vast majority of insights which shape our product vision come straight from calls with the customers.
The way we approach the actual software development for our product exceeds the scope of this article, but let me know in comments if you’re interested in discovering our tech stack and how we iterate.
Tools we use for product development:
- Metabase / SQL
Some of our tactics didn’t bring customers or leads. Or they didn’t do it in the timespan we were expecting. I’ll try to describe the reasons behind these failures but keep in mind this might well result from lack of experience.
Getting PR without having THE story
We don’t live in the car, nor have we left the comfy chairs in one of the massive consulting corps. We didn’t even get a penny from any VC. We didn’t join a startup incubator (didn’t apply either). We’re a regular, bootstrapped company that started as a software house, worked with successful startups as tech team support, noticed a niche for coupon software in one of the projects and decided to bridge the gap by assigning a part of the team. All I’m saying is that we don’t have a compelling story and the press doesn’t seem to be interested in covering it now. I think that’s why it makes little sense to hammer on every publication's door anymore. Let’s just wait for organic interest, sparked by our consecutive milestones.
Dev conference booth
We attended one of the biggest Polish dev conferences, Devoxx. We got a free sponsor booth for helping the organizers print QR codes. We thought we’d seize this opportunity to promote Voucherify among the 2500 developers. We prepared freebies and micro-contests to spark engagement. But it didn’t come. Devs were rarely interested in understanding Voucherify’s benefits. They were, however, curious about our job offers or gadgets. There were some exceptions, but not what we hoped for. Perhaps giving a talk would be a more effective way to grab attention.
We created 30 issues of Delivery Weekly - a newsletter for developers who deliver software in a fast-paced environment. Every issue had a collection of curated links about cloud providers, developer tools, product design, software architecture, etc. Although the effort put into curating the links wasn’t that great (our team shares such links on a Slack regularly), the number of subscribers hasn’t been impressive. We could invest more time in promoting or pivot into something more focused on a particular technology (like Cooper does), but we agreed to focus on SEO and product marketing.
Almost a year ago we launched a small cold-mail experiment. We sent out several hundred messages to marketers (Heads, VPs) but the performance was poor. Perhaps, the problem was the website back then or the fact that we had issues in selling the benefits in a short memo only. And we didn’t want to target developers as they get pissed at cold emails pretty quickly. But again, the small sample and the timing make us think it’s worth reviewing in future.
Similarly to cold emails, this channel was probably introduced too early. On top of that, we didn’t invest enough to see the results. Therefore, we ended up spending a couple of our first weeks on understanding the platform and the ad mechanics. If you add that our website was poorly ranked back then, it becomes clear that we just burnt money. The default settings of Adwords should be smarter so that early stage companies don’t waste their budgets.
But the fundamental thing is that most people ignore Google Adwords and go straight to organic results and I bet this ratio is even higher when it comes to tech people. So, as general advice I would give up on paid ads until you get ~100 customers organically and know what keywords will attract your target - and still I wouldn’t treat it as your main marketing channel.
Well, I wouldn’t exactly call it a failure, the thing is that we just don’t believe it translates into leads. The buzz around the event and therefore the attendance was just okayish - even despite the decent reward. And the low attendance resulted in fewer integrations/plugins/connectors to Voucherify than we hoped for when coming up with the idea. Anyways, we had great fun and took some valuable lessons from the Hack Night which I have described here.
Learn how to improve your customer acquisition & retention with Voucherify, all-in-one promotion platform
So, to sum this all up, our advice for an API-first SaaS marketing strategy:
- Listen to customers, keep the feedback loop warm all the time. Don’t run away from in-person communication (calls, f2f meetings)
- Make sure your product documentation is transparent, rich in examples, and interactive
- Invest in SEO early on
- Content is not only relevant on a blog, think of open source/community artifacts you can share
- Hold back paid ads until you start getting customers organically
- Experiment - your product is unique and as such should be your go-to-market strategy
- Patience - it’s not a sprint, it’s a marathon. Be patient. Like really patient. We’ve learned that the smallest time unit in big corps is a quarter, e.g. if somebody cannot join the 5-person meeting, the next possible date is in 3 months :)