How we got scalable

The predictable deliverance of online reviews, as explained by our CEO

The other day, my partner Jurjen came into my office. He was bouncing off the walls with excitement and asked me to check out the campaign results for one of our new clients. Intrigued, I opened our Mission Control Centre and looked at the review harvest we had reaped for them. A long, glorious list of stellar reviews appeared. Lengthy and detailed, these reviews had an average word count of 83 and an overall rating of 4.2. The product: old-fashioned toothpaste. The client? Unilever. We high fived. Our review machine had done it again. In the last year we proved that we could reliably deliver online reviews for a wide variety of products, from cars to sextoys, from band-aids to make up, sneakers to wetsuits, built-in ovens to enormous refrigerators, and from scissors to coffee machines. In 14 countries, in 20 languages on a multitude of online retailers.

Being able to deliver online reviews isn’t as easy as it looks. We call it an art. And hard work. Hard work art. In the starting phase of our young agency, we successfully delivered online reviews through pure dedication and a whole lot of overtime. We grew significantly in the size of our workforce to meet the ever-increasing requests of our clients. We used more than 34 different kinds of software on a daily basis to secure results. Google sheets, Mantis, Road Warrior, WordPress. But we were overwhelmed, with little in the way of oversight. Looking back at our start-up phase, I have to admit, it was a nightmare of systems and programmes gobbling us up with their sharp bureaucratic teeth.

Dreaming of a review machine

We had to change our approach or we wouldn’t make it. I read a book that changed the course of Stars and Stories and completely shifted my paradigm. Before Stars and Stories was born I wanted to be a Don Draper, a true Mad Man with an agency. And as handsome. After the book, I desired to be like Henry Ford. But not so ugly. I felt it was more important than anything else to create a machine that would take all the work out of our hands. The book is called The E-Myth Revisited and tells the tale of entrepreneurs trying to turn their passion into gold, only to be thwarted by their own success; when you attempt to handle your growth without a system, you only limit yourself, become overworked, stressed and deeply unhappy. The author’s message: build a machine to do the work for you.

Understanding our constraint

We wanted to find a way to get the job done for us and so asked ourselves: how can we replace the human labour of delivering reviews with software? How can we become a cookie-cutting manufacturer for online reviews? We started by describing step by step what we had to do. First we had to find a big white board. Then we boldly wrote: ‘6 steps to review success’ before adding detailed steps to each of the 6 pillars. We were shocked to realise that we had to check and pass more than 250 steps from briefing to completing a campaign. We also discovered we needed around 20 hours to prepare a campaign, 40 hours to run it and 20 hours to complete it. To make matters worse, we would utterly lose sight of campaigns if a campaign member went on a holiday, was sick for a couple of days or even worse, quit the job. We faced our constraint and looked it straight in the eye.

Adding a fresh perspective to the challenge

We were stuck in our sluggish method of campaign management, and decided to get an objective party to take a look at our quest for a fool-proof review system. It worked. Actually, on a side note, asking for an alternative view to one of our challenges, has always helped us tremendously. We hired a Vietnamese party to design our vision of a review machine, because at that time, offshore development was the only thing we could afford. Communication was difficult, but they came back with some really cool ideas and we used them to develop our machine further. We hired developers to start building our dream machine.

Don’t settle for anything less when you’ve got momentum

As a team, we realized we had to change fast in order to grow fast. We understood our restraints and had, in theory, mapped out a machine that would help us deliver online reviews in a scalable and predictable way. We used this momentum to whip each other into action and carry the whole team forward. I didn’t accept ‘no’ from my developers. The plan on paper had to be realized, so we prepared to work with the new system from the development deadline onwards.

Making the abstract accessible

The whole idea of the machine was, even during the development phase, still very conceptual. When you have to complete more than 250 steps to deliver a perfect review, it is hard to believe a machine can do this for you. We needed to make this concept tangible, so we decided to give the machine a name. The name of a person that reflected the spirit and the potential of the machine. We baptized it ‘Musk’ after one our 6 company heroes, Elon Musk (the others being Armstrong, the Wright Brothers, Rosa Parks, Einstein and Sherlock). Strangely enough, from the day we started referring to Musk, we all started to really understand it.

Testing our assumptions

We could program Musk to do whatever we wanted it to do, but also understood that completing 250 steps to get to a perfect review was uncool. We singled out the steps that took a lot of time, and thought outside the box. One of our time-consuming tasks was communicating details of the products up for testing with our reviewers via email. For our campaign managers, it was reasonably doable to get this info for one product but as soon as we had more than 3 products to test during a week, it got more complicated. From creating the email to accommodating revisions from the client, to A/B testing the mails prior to sending, getting more products to test meant that the complexity of the process grew exponentially. We decided to play with the idea of not using any text at all. What would happen if we only showed pictures in our emails? This would save a lot of time and we would improve in quality and speed. A/B testing the assumption showed us an increase in our reviewer click through rate with over 100%.

Scalable result

We are a 4 year old company and Musk was only integrated one and a half years ago. Our revenue exploded (we recently won an Award for fastest growing company in the Netherlands…) and so did our output. But simultaneously, we halved our team, opened business in 4 more countries, and whittled campaign management down to a couple of hours. Musk handles around 350 campaigns a month, and if we would triple this amount, we would only need one or two more people to get it done. And the best thing? Every single time we push the button in Musk, we get great authentic reviews from real people.