“Today CRM is Much More a Tech Game” – Interview with Steffen Harting
Steffen Harting is an experienced digital marketing manager from Berlin. Over 20 years, he led online marketing efforts for Sportfive, Accenture, Delivery Hero, Wundertax, and most recently Inkitt, a reader-powered publisher.
Check out more interviews:
- "My interest in tech came out of frustration" – Interview with Annika Rabenstein
- “Don’t be afraid of the technological side of marketing” – Interview with Nick Allen
What’s the first thing you’d say to a newcomer in the CRM space?
Today, CRM is much more a tech game than originally. So the first thing a CRM manager definitely needs to know is analytical thinking. I still see many people who aren’t analytical. In the past, a CRM manager was more focused on how to create an email copy, a more creative-minded person. Today, a good CRM specialist is an engineer or a data specialist who has a basic knowledge on how message templates might look like.
Let me repeat this, today CRM is a crazy tech game.
I guess that when you started your career in the digital marketing, technology wasn’t widespread. How did the marketing technology landscape evolve over the years? And how have you as a marketer adapted?
Yeah, getting fast feedback on your marketing campaign in 2000 was beyond the bounds of possibility (laughs). It took ages to understand if what you are doing is successful. Also, with respect to scaling it was really difficult and risky. Now you can get instant feedback in a few hours.
Things have changed around 2010. When I was working for Delivery Hero, when it was still considered a startup, we’ve seen an outburst of martech software. They came with great sales pitch decks, promising huge ROIs and whatnot. But when push came to shove, the deployment fell flat because of the mismatch between marketing and engineering teams.
And often it was not about the technology itself, many adoptions failed because of people issues. This is why it’s important for a CRM manager to understand how a tech team functions. I’d say that the number one thing to understand is the psychology of the tech people.
Agreed, we wrote an article on how to thrive as a marketer by befriending a “grumpy” developer.
Exactly this. Marketers think differently than engineers. Tech people often don’t know who a customer is, what they like and what they want. Marketing people, especially good marketing people think of it first. When these two teams talk to each other, there’s often a disconnection.
It manifests right in the beginning when marketers don’t even know how to brief a tech person. They can’t create precise requirements that the tech people expect. If they forget something, then developers will create a wrong thing because corner cases are missing.
Then, the marketing complains and developers become even more grumpy.
Yeah, exactly. Not understanding the tech leads to a wrong brief which leads to a wrong code — lots of waste. Another thing I also preach is that marketers should understand the limitations of the systems. For me, it was really a long trial and error lesson to understand that a campaign might fail because of some technical boundaries, like for example API limits.
The story I remember vividly was the time I wanted to integrate Zendesk and Typeform. While setting it up, I introduced a tiny — from the marketer’s perspective — misconfiguration. I was shocked to learn that you can break the system by changing the name of one little question. It didn’t impact customer experience but it did impact the backend.
I had similar stories again and again, this taught me a lot at the end of the day. Once I had to set up a website for my mother-in-law who was overpaying for website maintenance. I chose Joomla hosted on Apache server and I spent several long afternoons in front of the keyboard to get it done. Only to realize that it was available by the hosting provider with an auto-installer. I didn’t regret it though, as now I have a much better understanding of how these things work behind the scenes.
For instance, when I was asked to add several tracking software to the website, I ended up with small spaghetti code. I learned why I need a tag manager. This is obvious today and I always start with a tag manager, but after my lone hacking nights I understand the “why”.
Any other areas in marketing where tech skills are relevant today?
SEO. Without technical background, without understanding the basics of HTML, you can’t become a complete SEO manager. It is and has always been about the content of course, but in the heydays you could get a good SERP rankings without tech knowledge, just buying links.
And finally CRM. Successful campaigns are run by data-savvy people. It’s not so much about creative copy, it’s about targeting with the right message. One thing I’d recommend to CRM and performance marketing people is taking a course in SQL. This will help them understand data basics and so become more autonomous in your job. Today, data engineers in general are scarce. Let alone data engineers assigned to support the marketing team.
This will also teach you that on the higher level of data analytics, there are always issues. That building a data analytics pipeline is easier said than done. You’ll learn that you need to structure data, clean them up, and take care of its integrity. That the data you need are not in one simple table. That you need some computing power, you don’t always have at your fingertips, to update the data in an instant way.
I remember a funny story from Delivery Hero. We had a generic discounts, a fixed coupon code like “Summer2015”. One day, a lot of people used it and the whole website went down. The tech team complained to us about adding the code without consulting the engineering side, but the team had always done it this way and it had been fine. It was the scale which caused the fire. It turned out that every time a customer applied the code, its validity was checked against a database with millions of codes. Now, you can imagine what happened when thousands of people redeemed the code at the same time.
Yes, a lot of finger-pointing. Funnily enough, I sometimes tell the marketing team that they get an extra bonus for crashing the system on one condition — the campaign is so good it brings too much traffic in. Then, we get at least one good thing out of this (laugh).
So, SQL is one skill a CRM manager will find beneficial, what else?
There’s something more to SQL — statistics and interpreting data. One thing I’m seeing more and more is the inability to verify data. Junior marketers often don’t have the ability to realize that some data look awkward, especially if it proves their point. If the data doesn’t confirm their assumptions, they start to question things, but the outcome of an experiment is positive, they become less watchful.
Recently, we had an experiment which showed that one little UI change resulted in 200% uplift in conversion and the team was happy about it. I say: folks it’s not possible that changing a color can lead to such a boost, there must be some issue in tracking. Similar errors come from experiments that don’t follow a statistical procedure, like using insignificant samples.
You need to learn how to run sanity checks in your head to quickly evaluate the outcome of the campaign before you want to draw insights on top of it.
What else? Ideally, a scripting language like Python. I don’t know Python myself but I plan to learn it to automate more stuff myself. Automation is now the biggest trending in marketing in my opinion. Even the basic knowledge in automating things reduces the necessity of reaching out for help to external teams.
That’s right, today we can see more and more marketing APIs doing the heavy lifting for you. The only thing you need is a knowledge of a scripting language to leverage them.
How to get devs on your side then?
As I said earlier, the key is to understand how they operate. How they prioritize tasks and how they form requirements. If you brush up on this, it’ll pay off.
One tip is to involve the tech team into marketing software tools selection. If they take part in the assessment, they shouldn’t have the excuse later on (laughs). If you pick a tool as a marketer and devs have to use it, they will always complain. And you want to avoid this.
Psychology plays a big role here, so if you master “human API” and supplement it with basic tech knowledge, you can do wonders.
So, to sum up– working with data, learning what are the limitations of systems you work with, including web and mobile applications and 3rd party APIs, befriending devs by understanding their work and how software is produced in general.
Thank you, that’s a nice wrap-up! Any specific resources you can recommend to junior CRM specialists to get started in these areas?
I already mentioned an SQL course. What’s important here is not to stop after the course. You should create a sandbox where you can practice without worrying you’ll break something. There should be an easy-to-organize database to play with in every company. At least in the read-only mode, and not the whole database at once, but one or two specific tables to get started.
I also recommend books. The type of books is shifting now, though. Ten years ago, I suggested SEO-related books, because back then the core marketing activities went around Google search. Now, it has shifted to Facebook Ads, or programmatic advertising in general, especially in the app space. And I don’t even have a book to recommend here yet but I advocate “Hooked” by Nir Eyal and here’s why. We know that marketing can only be good if the product is good. The first task of marketing is to understand what’s the product, what makes it good, and align its own priorities respectively. I’ve seen many teams trying to jump into big time marketing before getting the product right and they failed. “Hooked” helps you realize just that.
Finally, I’m more and more inclined to recommend a general book explaining the basics of software, because with the speed the tech is entering the marketing world it’s becoming a must-read right now. I didn’t find any position so far (laugh). (MS: recommended “Code: The Hidden Language of Computer Hardware and Software” !)