Fostering a product culture in a large company
Developing a product culture in a large company is by no means trivial and the challenge is different depending on the company since each one has its own culture with its particularities.
In this article, I will go through how we developed a product culture in a large company with a strong data-driven culture, whose expertise is the optimization of acquisition costs and conversion rates in the digital channel of Fortune 500 companies.
Move fast and measure things
In such a context, due to the large quantity of data the partners´ websites generated (well-established products and brands), it was key to be able to read the data properly and get the right insights to optimize the funnel and grow the digital acquisition. The user itself was not approached and only quantitative data was considered in the decision-making process.
Building fast was also something to be proud of. As soon as the quantitative analyses showed the results, the team gathered the insights and built the new feature or test. That type of iteration led to the empowerment of the business and data teams over the design team, after all, the quantitative data was the main source of truth.
Although this line of work used to work well at other initiatives with a more tangible goal and next steps, it proved to be not the ideal approach to build a startup from scratch.
The first startup initiative
In 2018, iq was founded in the Brazilian office and became the first startup under the Red Ventures portfolio. Building a company from scratch required very different skills in comparison with those required to operate the acquisition funnel of a well-established brand.
The mistakes of the beginning
In the beginning, we tried to mirror the partnership´s operation in the startup´s, which led us to make the following mistakes:
1. Not talking to customers
This is a classic mistake that many startups under big corporations fall into. At iq we took more than 6 months to start talking to customers. That is linked with the company´s culture and way of operating. Talking to customers was not a habit as well as not perceived as valuable.
Talking to customers was something so far from our culture, that we delegated it to an outsourced call center, which does customer success for us. People who had not a great understanding of the product were the ones selected to do this job.
2. Trusting only in quantitative data
As you can imagine, a young startup has almost no data to rely on. So, the data-driven approach had almost no use at the beginning, shifting the source of truth to an even more dangerous place: intuition and opinions.
3. Rushing to deliver
Testing was done live on the website. Our amazing development team is very fast and has gotten us spoiled since in less than a month we could deploy a new complex feature. Although the quickness was great, testing in production is very costly and can become demotivating. It was not uncommon to see huge releases having no impact on customer retention or acquisition.
How the change began
To start changing the way we built products we took the following steps:
- Talking to customers
Gradually, we started with 3 to 4 talks every 15 days until we got the habit of talking to customers every week for at least one hour, which is still not ideal. As I view product management, if you are not close to your customer, it is impossible to perform your job at a high level. So, my focus was to create a habit and gain knowledge from how our customers were using our product and what pains they had.
Talking to customers can be done by only one person and it was our starting point to develop a product culture. The book Talking to Humans helped me improve my technique on customer development and I highly recommend it.
2. Sharing customers feedback
The second step taken was to share customer quotes and audio recordings. It was great to see how people reacted when they faced users interacting with the product in ways that we had not anticipated. Sharing this information was crucial to start getting buy-in from the senior people in the team.
3. Driving results
Talking to customers is great, but it is not enough. Results are what make people truly change their minds. So, by talking to customers, we were able to identify why our referral program was not escalating.
The reason, discovered in customer calls, was simply because people were not aware of the referral program. After correcting it, the results spiked reaching a plateau. Again, we interviewed some customers via phone and discovered that they referred the product but were not aware whether their friends had joined. As soon as we fixed this issue, the referral rate spiked again. Talking to customers and focusing on the root problem paid off, and we increased our referral rate by five times.
Taking the next steps
After the first success, everybody understood the importance of talking to customers and how it was changing the quality of our releases. However, customer development and product discovery are much more than calling customers, requiring more creativity and a set of different approaches depending on what you are trying to uncover.
To continue improving our way of building products, we did the following:
Empowering design: the design team did not participate in the decisions of what should be tested or built next. They normally only executed what the business team thought was best. We shifted that, partnering with designers and collaborating with them along the entire process.
Usability testing: After a lot of effort to get buy-in, we did our first usability test with targeted users and that was instrumental to advance our way of doing product. By showcasing the results of the test, we were able to go even further and not only do more iterations but also set the standard to test without code as much as possible before handing over to delivery.
The benefits of the new dynamic:
Decisions based on evidence: testing without code is the best way to avoid opinionated discussions. Even though we had not statistically significant data, we had evidence — data from 3–5 sessions with users — and that was clearly more probable to represent the truth than opinions.
Better results: during usability testing and other kinds of discovery practices we learned so much about how customers interacted with the feature in question that, when we shipped it, there was almost no surprise, customers were interacting the same way they did in the usability testing sessions. That means we became one order of magnitude faster since the larger part of our learning came from prototypes, which take 1 day to craft, rather than the coded website, which takes 1 month.
Current product-discovery cycle
Currently, each step is highly collaborative and non-linear. As we progress and learn, we go back and forth if necessary. With this process, we are able to unlock a ton of creativity and good ideas from our designers and developers.
By questioning the way things were done, seeking to learn how the best companies were doing product, and starting with small initiatives, we were able, in a short period of time, to change completely how product is done in our company.
The key in this process is to be bold and push forward (sometimes without asking for permission) and communicate to everybody why and what you did, along with the results. We still have a long journey ahead to improve and spread the product culture.