International speakersPrestigious award showThe latest insightsAnniversary edition in the Netherlands

Martin Röttgerding on “Smart” Shopping and smarter use of Ad customizers

09 January 2020

Martin is Head of SEA at online marketing agency Bloofusion Germany. As a PPC geek, he loves to take things a bit further – by digging into the mechanisms of Google Ads, building tools, or doing cool things with data. He will be presenting his latest work at Friends of Search on the 4th and the 5th of February in Amsterdam and Brussels.

How did you get started in Search Engine Marketing?
I have been in search engine marketing for 15 years now. I started as an intern at Bloofusion, where I worked for Markus Hövener doing both SEO and PPC. The internship turned into a permanent position and I basically became employee #2. Around 2010 we agreed that I would shift my focus towards PPC and develop this area for the agency. I actually had mixed feelings about turning my back on SEO, which I found more interesting at that time.
After a while I started researching Quality Score and made a somewhat huge discovery: QS is basically click-through probability. All the talk about quality, user happiness and so on was just marketing. I wrote a series of blog posts called “The Secret of Quality Score” to present my findings. It didn’t generate the buzz I thought it deserved, with many people skeptical and experts dismissing the conclusions (it’s all common knowledge nowadays). Still, it gave me a lot of confidence and showed me that I have something to contribute to the industry.

What do you think is the biggest difference after working in both PPC and SEO?
SEO has a reputation of being something magical, or an art, while PPC is more about numbers, machines and automation. SEO is often about the coolest strategies to get people to link to and talk about the content being produced and to rank it in search engines. PPC is important for performance, but compared to SEO, there are few secret creative strategies. In my opinion there are many talented people in the PPC industry, but not many people blog or talk about what they do. Getting feedback is also a lot more difficult than in the SEO community. Not because the content is secret, but it just isn’t part of the PPC DNA.

You are very innovative and have come up with a wide range of creative solutions in PPC. What is your biggest success?
That is certainly the search term segmentation strategy for shopping campaigns. The strategy is about using keywords where Google didn’t intend for them to be used by constructing a double negative of sorts. As a result, you can segment your traffic based on its value and bid accordingly – something that Google never intended to be possible.
That strategy put me on the map five years ago. It was a combination of finding the right thing at the right time and having a great platform to spread the word. I spoke about this at Marketing Festival and they let my post a video of the presentation on my blog. It started slowly, but people wrote about it and talked about it on other conferences, often giving credit to me. Special thanks for this goes to Kirk Williams, who has been a big proponent of the strategy.

Where can you find the inspiration for these things?
I think my inspiration comes from very often looking at a problem from different angles. I don’t really have a certain process; ideas come if you think about something long enough. Some things such as the segmentation strategy for shopping campaigns are about simple logic. I have the luxury of not having to work on customer projects every day, so I can take some time to mull over problems and come up with new ideas.
That said, failure is definitely part of the process. For every cool idea that I can eventually talk about, there are many others that don’t make it that far. I try out a lot, especially with tools. Some things don’t work at all, some are too specific, some aren’t interesting enough. The best ones make it to conferences… you’ll see ?

Your approach with campaign priorities for Google Shopping was introduced five years ago. Has it developed further since then and what are the most important changes?
I think the strategy has been fairly stable over the years. It is still about dividing shopping campaigns into two or three layers. The distinction between brand and non-brand remains key. There’s some additional content about troubleshooting and large campaigns on my blog, but not really any changes.
We’ve done this for a lot of customers, so naturally we’ve made refinements and build our own set of tools to handle things a bit more smoothly, but at the core it’s still the same idea.
Sometimes it makes sense to deviate from the standard model and use other segmentations, but these are all individual cases. The beauty of the original model is and always has been that it works well for everyone.
It’s unclear what the future will bring for the approach. Perhaps it will one day become obsolete because smart bidding already takes the different levels into account. Maybe people will switch completely to smart shopping campaigns and give up all control – who knows?

Google pushes many of their smart solutions, what is your vision on this topic?
I think “smart” should be in quotation marks. “Smart” means that Google does things for you. The message is of course that Google’s systems will do a great job, with buzzwords like “machine learning” everywhere. In reality, these “smart” solutions are designed as black boxes. This means that all you can do to assess their performance is look at the KPI’s that Google provides – that’s not good.
For Google, these solutions make sense. The system grows and gets more and more complex, with new channels, formats and other things added all the time. “Smart” solutions provide a way to combat this complexity. The last thing that Google wants is for people to stay away from the system because it’s too complicated. It’s important to realize that some “smart” solutions are aimed at people who want to do something but don’t have the resources to do it properly.
Personally, I hope that Google will keep both options open: The “smart” solutions for beginners and the more complex approaches for sophisticated advertisers. However, I’m unsure whether this is where we’re going. Take local campaigns for example: Even though it’s not in the name, they are basically a “smart” campaign type – a black box with hardly anything you can control manually. Another bad sign: Google is pushing advertisers with regular shopping campaigns to use Smart Shopping instead. I think this is a very bad idea.

What are you currently working on?
These days, I spend most of my time in tool development. My current focus is tapping into the Google ecosystem. It’s too early to go into specifics, but the idea is this: For us as an agency, structured data is very valuable. However, asking each client to provide data specific to our needs is impractical. Yet they all do it for Google. So why not build on this?
The potential is especially big for retailers. A product data feed alone holds a lot of potential, but there’s no need to stop there. Google Analytics has some performance data on the same products. Google Ads has data from yet another perspective. And that’s just the beginning.
One important aspect is that things need to work within existing structures – you don’t want to throw away your campaigns in order to use a new tool. One way to accomplish this is using ad customizers. In this area, we came up with a new approach that I will present at Friends of Search.

What can people expect from your session at Friends of Search?
My session will be about ad customization and how we can make it more accessible for everyone. Ad customizers are one of these great features that people rarely use since it’s a bit more complicated due to the involvement of data feeds, buggy interfaces, and poor documentation.
To counter this, we have developed a simpler approach to ad customization. At the core it’s this one weird logic trick that apparently no one has found before. You could call it a hack… it basically comes down to preparing some things once and then never having to deal with data feeds again. This is especially helpful for teams, where only one person needs to do this and then the whole team can reap the benefits.
In my presentation I will show how my approach for ad customizers works and then give you some ideas about what you can do with it. Some will be simple; others will be really advanced – up to ad adjustment based on the weather.

Unfortunately, we have decided not to organize a Belgian edition of Friends of Search in 2024. The 11th edition of Friends of Search we are organising in the Netherlands. Belgian visitors are of course more than welcome. Join us on March 21, 2024, in de Kromhouthal in Amsterdam.

Go to Dutch website Stay here