What We Learned Building Just Ads

I decided to build Just Ads five years ago based on two predictions after encountering the mindless drudgery involved in performance marketing:?

• In the future, marketers would interface with software that allows them to upload creative assets, refine target audience, manage budgets and KPIs, review performance, run experiments, and make decisions. The launch of responsive ads across platforms validated this insight.?
• Machine learning APIs would allow marketers to plug in their data and derive insights with minimal investment in research, aiding the aforementioned use-cases. The launch of OpenAI's GPT-3 API (and DALL-E which was used to generate the cover photo) validated this insight and we were quick to integrate this seamlessly into the product.

Ad platforms such as Google, Facebook and TikTok require roughly the same inputs (creative, targeting, budgets, goals) but with a layer of abstraction (campaigns, ad sets, etc) that varies across platforms. This second layer is where agencies and consultants apply their expertise and best practices. Not only does that create a tonne of manual work though, it requires separate expertise for each channel, expertise that is short in supply and hard to gauge given how easy it is to secure certifications offered by the platforms themselves. This makes it very difficult for most businesses to advertise effectively across the various stages of the customer journey without a significant investment.?

To make this easy to grasp, we came up with an analogy. Imagine you had to build a machine to automate cooking. You’d need to design a user interface that allows anyone to provide and store ingredients, and a back-end that stores recipes by experts along with the utensils required to execute the instructions. In our cases, creative assets, targeting, budgets and goals are the ingredients. API connections are the utensils. Meta templates are the recipes that dictate how ingredients are turned into campaigns behind the scenes.?

We took it as our job to absorb the complexity and make the process easy for the user, and built a great first version of the product (video walkthrough below).?

Yet, we failed for now because we did not foresee other problems.?

1. The scope was just too broad given our scarce bootstrapped resources.
2. A global pandemic, impending crises, and sudden loss of revenue caught us off guard.?
3. We could not launch till we satisfied the extensive minimum functionality requirements imposed by Google and Facebook. This imposes a cost on building disruptive products in this industry without seed funding, which we failed to secure.
4. With service revenue funding product development, complacency set in as we assumed we would be able to raise seed funding at some point to build the functionality required for launch. This could have been overcome by having a minimum viable product that could skirt these requirements and generate revenue for the SaaS product independent from ad APIs.?
5. Moving to New York isolated me from the teams in both Stockholm and Lahore. This made it difficult to shape organizational culture across three continents, develop effective processes for both the engineering and account management teams, and drive sales.?
6. Entrepreneurship is harder for immigrants. I chose to uproot my life in Sweden and move to the US because I did not my wife to be another Pakistani woman who got a prestigious medical degree only to abandon her dreams. More than five years later, I am not allowed to even apply for work authorization on my H4 visa, which could be a blessing in disguise forcing me to stay on the entrepreneurial path.

Now, it's back to the drawing board with a wealth of new insights, lessons, skills, tools and opportunities. Performance marketing remains drudgerous, our vision makes sense still, and there is ample room to build in a way that empowers marketers with machine learning fueled delight.?

Yet, pivots don't have to be narrow. Slack made video games before they pivoted to building a workplace collaboration tool. Netflix did video rental before pivoting to online streaming. Whatever we do next though, we'll be applying everything we have learned, and do a better job at building products that make the world a better place.