logo UE Guvernul Romaniei Instrumente Structurale

Welcome to Growth Interviews!

Welcome to Growth Interviews, the fun, stimulating and engaging series of conversations driven by digital business growth.

Our mission is to provide insights and ideas from world-class professionals on the topic of growth and to cut through the noise of so-called marketing tips and tricks, revealing the money-making strategies behind e-commerce.

Each episode is an intriguing challenge involving an insightful expert who reveals some of their best-kept secrets, which you can use right away to boost your business. 

In this week’s episode of Growth Interviews, we invite you to join our conversation with Brad Geddes, co-founder at Adalysis and author of Advanced Google AdWords. Brad doesn’t refer to himself as an expert or guru, but trains marketers and lets the results speak for themselves. He is, however, able to flawlessly teach and practice SEO, PPC, CRO, affiliate marketing, product development, product positioning and much more.

Brad’s agenda is always packed when it comes to creating workshops in Advanced Google and Bing Ads, AI and machine learning. His official website is proof of that.

From his long experience in e-commerce and digital marketing, he shared with us some amazing ideas on how machine learning can become your best friend, especially in PPC campaigns, or in short, AI PPC management.

 Also, be ready to find out some exciting 2019 e-commerce tips.

Check out the full video interview below, where all you need is 8 minutes!


Brad Geddes gave us such valuable information, that we think it’s high time we looked at the written words and keep in mind his topmost interesting ideas.

The spiral of new beginnings in digital marketing

How did you first get into the digital marketing landscape?

Brad: That was over twenty years ago. I initially started with just designing websites in all these different industries. I quit my job and started affiliate marketing. I knew how to make websites, so I started doing organic and then soon after that the first PPC engine: goto.com launched. I called them up to make an account and it sort of spiraled from there.

What frustrates you in the digital marketing landscape?

Brad: People not knowing how to work properly with machine learning. Often it’s more like there’s a battle: here is machine learning, here is us and people are often fighting with it. It’s not that the machine learning is taking anybody’s jobs or anything, it’s that with all technology advances, we need to learn how to leverage it properly. We’ve got the good creative side, the strategy, and then the machine learning is doing a lot of the heavy lifting. But it can’t think like a human does. So, it’s really about people battling machines as opposed to working properly with them strategically.

Everyone has heard at least once about machine learning. What does it do exactly? It begins with observations or massive quantities of data, finds patterns and makes better decisions in the future based on the data users provide. The main purpose is for computers to learn and adjust without the need for human assistance.

At Omniconvert, there will be a future launching of an experimental software product, developed with the support of the EU non-reimbursable funding, through the Competitiveness Operational Program (2014-2020). This product uses a machine learning algorithm that will allow online stores to launch interactions with visitors based on their real-time behavior.

The road to success in e-commerce marketing

Tell us a good story you have from working with one of your customers.

Brad: Often what you’re looking at are people who are not using surface tactics, but take that and do deep dive with the tactics and how they work.

We have a rather large B2B company as a customer. I used targeting – it’s not new. It started five-six years ago on the search side. They took it and said, ‘All right! What leads do we really need? We need to know everyone who subscribed to each one of our individual products. We need to know who is sharing our content., who has downloaded whitepapers, but has not become a subscriber, who subscribed, but less stayed in. In SAS you get hockey sticks where someone sort of goes on and boom! their pucks go flying. So who has hockey sticks and who has not.

They created roughly 300 audiences and then layered them all in different strategies, throughout their marketing. So, they do an amazing job where most people would just have, ‘Here’s who came on the website, here’s who bought, here’s who didn’t, let me do some remarketing and be done with it.’ Really thinking through what is your full funnel strategy and building audiences for each level of that can link towards some really fun marketing instant results.

Top 2019 tips and tricks for e-commerce marketers

Tell us what do you think that e-commerce marketers are missing out these days.

Brad: A lot of it is using some of the new affinity and in-market audiences across the Display Network. When you look to see what you got, you drive people with search, social or organic and you make your audiences from often a remarketing – dynamic remarketing – standpoint. What you’re not seeing are people who are researching the same products that you have, but just haven’t been to your website yet, so they’re not in your audience list. Somebody in-market, the YouTube remarketing, the YouTube market list and the affinity list lets you reach people who are in that same research, similar to your remarketing list except they haven’t been to your site yet. So, it’s a good way of expanding where your sales prospects are coming from.

What would be some tips that you would give for e-commerce marketers?

Brad: Number one: not enough to good ad testing. In e-commerce you have huge amounts of ads, so no one really wants a test in an ad group. Looking at pattern-based testing or multi-ad group testing to get better customer insights from their ad data, that can then be leveraged into title tags and email and everything else, is super useful.

Number two: if you’re e-commerce, then obviously you’re running shopping campaigns, so be smart with your campaign segmentations. Ensure that you are running brand, non-brand, third party stuff correctly with your priority list and segmenting out parts correctly, so you can manage the margins across them. Of course, on e-commerce, you know the average person takes roughly four visits to convert, but it depends on which type it is: apparel is eleven and travel is twenty plus.

So, ensure you’re using your audience not just to initially create years, but also bring everyone back for the eventual conversion and then post-conversion: what does the upsell with the audience look like, as well, and segment your audience list so you can do proper up-sells and cross-sells based on what someone bought on the website.

E-commerce marketers are always on the lookout for all the complex details that make a PPC campaign successful. Above all that, you should never forget about the basics in ad campaigns, since they are the very first lesson you had when you started online advertising.

A hard piece of evidence is this article we compiled a year ago when we felt that digital marketing altogether was missing something crucial: the receiver itself, or in our case: the customer. Therefore, we had to go back to square one and revisit the basics of communication and redefine our path from there. The same is true for every element of digital marketing!

Tell us your take on customer retention.

Brad: That’s a great question. Are you in an industry where your buyers are repeat-quickly or repeat-long term?

If you sell cars, no one’s buying a new car every month, or at least if they aren’t kind of jealous, so that’s a brand strategy of a long term, where you’re dealing with support issues and stuff for years until they buy their next new one. The automotive industry is a terrible job, of course, because it’s so hard to do long term. Whereas, if you got people who are rebuying stuff on a very regular basis, you need to rebuy a lot of food supplies, if you have kids, you’re buying a lot of the same supplies month over month.

Some people do that quite well, others are very lazy about it. It’s generally done via email and not really thought of as, ‘These are customers. They’ve proven they’re going to be a customer. We know that they reorder diapers every three months, they reorder baby formula every two weeks, they reorder new school supplies once per year.’ Run some remarketing off of when subscriptions renew or when things renew to ensure that you’re keeping in mind awareness on top of your loyalty things. A lot of emails are never opened. If you only rely on a single channel on customer retention, you’re missing out, compared to reinforcing that in other places.

Automation makes room for human creativity

Tell us what excites you these days.

Brad: I get excited by fun large scale things that work for a lot of people. These days an audience is an easy choice. It’s gotten down to: Google, Bing’s and even Facebook’s APIs are very fast to be able to bring in data. Use your own machine learning, or just filter those, and push things out again. It’s about the level of automation we can do to get rid of the stuff you should not be doing. No one should be doing week over week or month over month reports of the same thing. They are a complete waste of time. Most people should not be doing bidding.

Considering the way we can now move data around, bring it locally or just use automation that’s already in the system, we can spend as humans more time on creativity – the hot, stretching the fun stuff that gets you into marketing – and automate all the really boring stuff that no one wants to do anyway. That’s the fun stuff these days, getting rid of all the bad stuff, so you get to concentrate on what’s really interesting.

While there are some voices that say creativity is already being automated, there are also studies that prove people understand customers better than machines. Solving complex problems with variables such as a client’s budget or customers’ emotional drives is still well managed by qualified marketers.

There is no need to fear or replace AI! Instead, use it to increase human creativity.


It is clear that machine learning will never replace a good marketer’s job. More than that, using automation can create more space for creativity to flourish and can give you back the time you used to spend on tedious tasks. PPC automation is the next best thing.

Now is definitely the time to embrace the benefits of machine learning!

How did you find this experience? Was it insightful?
Spread the knowledge! ⤵️

proiect cofinantat UE