Our Tesla article “Tesla replaces Salesforce with a self-programmed CRM system” literally went through the roof. Scott Brinker gave me another tip. Not only Tesla, but also the rather unknown competitor of Uber-Taxi, the company Lyft, is building its own marketing automation solution.
Lyft, a competitor of Uber, is also programming its own marketing automation solution. Lyft currently only operates in North America (USA and Canada). According to Wikipedia, Lyft was founded in 2012 as a mobility service provider, achieved a turnover of over 2 billion dollars in 2018 and this represents a market share of approximately 39%. Uber brought it in over 60 countries in 2018 to over 11 billion dollars in revenue.
A sting in the wasps’ nest? A shaking wake or a look into unknown worlds?
Many reactions to the Tesla article were “We’ve done this before – individual software is out”. Both consultants and software providers agreed with this tenor. Nevertheless, there were also some vendor calls according to the motto “We also observe that more and more customers also want to program themselves and thus extremely individualize”.
That is also logical. If you put the abbreviation CRM in your mouth, you automatically have personalization and individualization in context. Mostly, this refers to communication, advertising and the offer, but of course hundreds of thousands of Euros are also spent on customizing. Which is nothing else.
So it is quite understandable that companies that make “make or buy” considerations calculate: “What does customizing cost me and what does it bring,” and ask the question: “How much effort does an individual solution cost me,” and: “What percentage of a solution is used on average, and how much idle costs (modules and functions that are not used) arise?
The Lyft engineering team had also gone through these thoughts and came to a decision. The team is building its own marketing automation platform. The “why and how” was published by the Lyft engineering team on the portal medium “Building Lyft’s Marketing Automation Platform”.
What was the motivation for a custom marketing automation solution?
“Due to the challenges of managing a marketplace, we needed to tightly integrate our marketing automation system with the most important internal systems – namely Customer Lifetime Value Prediction, Attribution and Service Level Forecasting in the marketplace. We needed to know exactly how much we needed to spend today and what the impact would be on the market 2, 8 or 16 weeks in the future.
When evaluating commercially available solutions, we came to the conclusion The most expensive components of the platform are the internal services – LTV, attribution and forecasting. The actual tendering interfaces to the advertising platforms were relatively inexpensive.
Since the creation and maintenance of the tender interfaces was less costly and the control over all our data enabled us to improve the expensive core components of the platform more effectively, we decided to implement all this in-house.
In retrospect, this was definitely the right decision – we made numerous improvements to the core components of the platform that resulted in significant cost savings, improvements that would have been much more difficult to achieve if we did not have the complete data pipeline to publish bids”.
Besides Tesla and Lyft, many companies are working on individual software solutions
The article by Scott Brinker shows that there are quite a few companies that do not want to get involved with standard software. But, and this is also our impression, Scott writes: “Although we know that there are many self-developed MarTech systems in the wilderness, it is rare that they are described publicly.” (quote from DeepL translated from the American)
What is the conclusion now? Which is better?
30 years ago I started my professional life at Yves Rocher and I experienced a 100% individual solution there. It was sophisticated in every detail. So I can understand this tendency or retro trend. What is now right? It is clear that there is not only black and white. In between there are – as always – many different shades of grey.
Every company must ask itself these questions, among others: What does my future software landscape look like? What is the best mix of standard and individual software? How much effort does a company have to make to adapt the standard software to its own needs? Where does this path become too expensive? Do you only make your competitors smart because they use the same standard software and sooner or later the customizing expenses will end up with them? Where does an individual solution perfectly support the own USP?
The hasty judgement that an individual (CRM) software solution costs significantly more than an “off-the-shelf” solution is rarely tenable in our estimation. If you look at the costs for additional programming and customizing over the years, you will certainly have your doubts. Not to mention the dependence on product and service provider. It remains a question of philosophy.
Note: This is a machine translation. It is neither 100% complete nor 100% correct. We can therefore not guarantee the result.