The 3 Key Principles of Growing Startups
Launching a startup within a large corporation may seem easier than testing the niche and building a brand-new product from scratch.
However, internal startups still have particular challenges and pitfalls that may lead to failure.
Petr Garmashov, the Head of Products at PropellerAds, a part of AdTech Holding, shared his experience exploring new niches within a large business and horizontal scaling based on the available expertise.
PropellerAds side projects overview
PropellerAds, one of the AdTech Holding products, is a leading advertising network in the Tier 3 niche. It has recently launched five startups related to digital marketing. Four of them turned out to be successful, while the very first one was closed within a year.
Here is what they did:
|Trads||An app connecting micro-influencers on Instagram with advertisers||Fail|
|Zeydoo||An affiliate network with private CPA offers from direct advertisers||Success|
|Notix||In-app and web push notifications service||Success|
|ProPush||A platform for monetizing traffic with push subscriptions||Success|
|Adex||Anti-ad fraud tool||Success|
Before we get to a more detailed view of some of these projects and their key points, it’s essential to mention the possible reasons why a business might decide to scale. According to Petr, three significant points indicate it’s the right time to try launching something new:
- The core business has a smoothly operating tech stack;
- All the processes within all teams are well-tuned and efficient;
- The company feels financially comfortable investing in new projects.
Some may assume that horizontal scaling might be a way out when the core company shows deceleration. However, Petr argues that:
‘I suppose it’s way too late to launch startups when the business is slowing down. In this case, it’s better to focus on the core processes rather than horizontal scaling.
PropellerAds Scaling Cases: Failed and Successful Internal Startups
The horizontal scaling of PropellerAds began with two startups launched almost simultaneously. Let’s briefly overview each of them and then compare both to draw some conclusions from this experience.
Trads was created as a digital marketplace app connecting Instagram micro-influencers and advertisers. The principle of its work is the following:
- An advertiser publishes an offer or a product they want to promote via a third-party Instagram blog;
- A blogger gets access to a list of advertising offers and posts an Instagram story promoting an offer of their choice.
The AdTech Holding assembled a team consisting of four developers and one sales manager to proceed with this project. The app exclusively targeted the Russian market and did not imply creating any MVP. In other words, the team did not focus on the customer development process, using the testing in the production approach.
Micro-influencers highly appreciated the app’s concept as it allowed them to access an automated advertising network bringing almost instant income. However, the process was not that smooth from the advertisers’ side.
The major pitfall was the brand safety issue that forced an advertiser to check the audience quality and approve every blogger manually. It was destroying the whole concept of automated promotion via Instagram and thus discouraged advertisers from being involved.
After a year, the holding admitted the project was commercially unfeasible and was frozen.
The principles behind Zeydoo‘s creation were perfectly opposite. It all started with the MVP, which did not require a single cent to invest in. All that the team had in the very beginning was a Google Sheets table with the list of CPA offers from advertisers already working with PropellerAds. Here is what it looked like:
Unlike Trads, Zeydoo proved to have all chances to bring profits before the development stage. What is also essential is that Zeydoo’s initial team consisted of sales managers and did not involve any developers.
The managers worked with the PropellerAds customer base, which significantly simplified the first sales, and the development started after it became clear that the MVP proved its potential.
The product development process was based on the customers’ needs and not vice-versa like in the previous case. Zeydoo technologies were mainly based on the existing stack used for the PropellerAds network.
Long story short, Zeydoo became profitable in 1,5 years after launching and split up from PropellerAds. It now has two offices, with the headquarters in Cyprus, and keeps developing as a separate brand.
Trads vs. Zeydoo: the Key Approach Differences
So, what made Zeydoo successful and prevented Trads from flourishing as well? Here is a brief summary of what turned out to be crucial for efficient scaling:
|Hypothesis testing||In production||Via sales without development|
|The first customers coming from the existing company||No||Yes|
|Using the existing tech stack||No||Yes|
These four points became a base for the next AdTech projects. In Zeydoo’s example, PropellerAds managed to create three more startups. Two of them started showing growth within the very first year. In the following paragraphs, we will consider the main lessons learned and developed into a working approach for horizontal scaling in the AdTech industry.
Key Points to Consider when Launching an AdTech Startup as a Niche Leader
Based on its own experience, AdTech Holding created its own principles of product planning. They touch all the stages of creating a product, from the initial idea to the release and further progress. Here are the key points that every company employee pays much attention to when launching startups.
#1 Generating ideas
The main principle here is that every team can suggest their own ideas and become a product owner/startup CEO. To collect and validate the ideas, the holding created a dedicated POC committee that has already got twenty-two ideas, eleven working groups, and seven MVPs within the first year.
The company stimulates collegues to share their ideas at hackathons, brainstorm sessions, and an open idea backlog. To validate these ideas, the company uses classical methods: customer studies, market research, competitor analysis, cost estimation, and MVPs.
#2 Keeping a Step-away From the Core Business
One of the things Petr emphasized was the principle of keeping a new project ‘a step-away from the core business. It allows teams to apply successful techniques and well-known technologies and hire people with the right expertise from the already existing staff pool.
Besides, such an approach gives access to the familiar customer database, which crucially multiplies the chances of success.
#3 Building the Right Team Focused on the Right Things
Every new team created for working on a new product consists of 4-8 people. Here is an already traditional team structure:
- Business Owner;
- Product Owner;
- Sales Managers (3-4);
The team’s primary focus is sales and product, not development. The head office of the holding is also involved in the process, but all team sprints are independent.
#4 Testing an MVP
Creating an MVP for testing and market research has become a golden rule for AdTech startups. Zeydoo’s MVP success is a pretty unique example that is hard to repeat: it’s not a usual story when a totally free-of-charge Google Sheets file can work out so fine.
However, MVPs remain one of the core things, although they usually take more time and effort than the ones for Zeydoo. Testing in production used with the Trads app showed poor efficiency in the end, so it became a no-go practice.
#5 Independent Tech Stack
Although a ready-working tech stack is a big advantage for a new project, it became clear that using it contains significant pitfalls. Using the core company’s technology maintained by its development team substantially affects delivery speeds.
Quick delivery is essential for building startups. However, dev teams working on the existing products don’t usually consider tasks related to the newborn projects as a priority. It significantly slows down all processes that meanwhile need extremely high speeds.
Zeydoo turned out to be a great lesson for us that changed the rules on launching startups and led to a global transformation of the whole holding infrastructure.
Now, AdTech offers convenient ready technical solutions for startups, but all systems are rearranged in a way that allows new products to stay technically independent.
Successful sales are the jewel in the crown of a startup’s growth, and having a large holding behind a new product boosts them a lot. Here are the main advantages that AdTech Holding startups get from the head company:
- Leads come from the existing customer database;
- Cross sales opportunities and cross-in-app promo (for instance, Zeydoo can place an ad on the PropellerAds website);
- Conference efficiency: one representative can promote several projects at a conference.
#7 Tracking the ROI
To make tracking startups easier, teams decided on quarterly plans. They help to understand if the startup is gaining ground or not. The key measure of performance before break-even is profit efficiency.
After the startup has reached the breakeven point, teams may start paying more attention to other metrics like revenue, customer database growth, market share, etc.
How does a Core Company Control Startups
Although startups work almost independently, and it becomes crucial for the technical side of the product, the holding management still has particular requirements for every new team. Here are the guidelines that startups must follow:
- Use the financial infrastructure of the holding: payment gateways, accounting department, analytics;
- Follow the Know Your Customer Policy to reduce the risk of fraud;
- Download key metrics in the holding’s general statistics to stay transparent for all stakeholders.
To Sum Up
PropellerAds’s experience helped me work out three golden rules for new ventures. Creating a startup within a step from the core business, keeping it technically independent, and focusing on sales and marketing are the keys to success, Petr says.
This ever-working path from the idea and its validation to evolving into a separate business has almost automated the process of creating new products, making horizontal scaling a transparent and inspiring thing.