Interview with Byron Marr, Head Of Paid Media at Aira

Our research continues! We have a great opportunity to meet talented, influential and inspiring experts who are true examples to follow. Here’s our next interview with Byron Marr, Head of Paid Media at digital marketing agency Aira.

Aira is a UK based full-service digital agency who helps medium to large businesses grow their customer base and one of the few UK agencies to be both a Google Premier Partner and Facebook Marketing Partner.

Byron has been working in paid media in various forms for around 8+ years. At the agency he works on strategy through to implementation, using multi-channel campaigns to drive lead generation and scale revenue for Aira’s clients.

Byron has spoken at various conferences such as BrightonSEO, The Paid Social Show, SMX Munich, on the topics of Facebook Advertising and multi-channel.

In 2020 he was voted in The Top 50 Most Influential PPC Experts and launched a paid media focussed webinar series Coffee And Conversions.

As always, here are the main topics for navigation:

Being totally honest, it was an accident.

I taught myself how to use Facebook ads and Google ads for YouTube a long time ago to promote some of the music projects I was doing and I found I really enjoyed it.

What made me want to make it my main career path was the mix that you get of getting to be creative alongside being commercially/data focussed. I love to learn new stuff all the time and the platforms are ever changing so it aligns very well to my personality type in that respect.

I’m also very competitive. I love the instant feedback either positive or negative of working in paid media. We’re all competing against each other in the auctions and also competing against ourselves yesterday to come up with new strategies and ideas to maintain and scale performance.

You need to do more than “PPC”.

Machine learning has taken away a lot of the time consuming legacy tasks that the average PPC specialist would do meaning you need to focus on other areas — where you can add huge value.

I would say the first is conversion tracking; accurate measurement and being able to feed in the right data is essential for effective performance with ML such as Smart Bidding. Learning how to use Tag Manager will enable you to feed in more data, which in turn, will allow you to make more informed decisions and create more audiences.

The 2nd is CRO, what happens post-click has a huge impact on campaign side performance. From a technical perspective, understanding components which may be causing friction, for example, a poorly placed form or a CTA button which doesn’t stand out on the page will provide immense value to your PPC campaigns and also the client’s wider business generally.

From a broader perspective, understanding the value proposition and what makes the particular product or service different or better than their competitors will not only improve your conversion rate, it’s also learning you can bring back into your ad copy and extensions to make sure you stand out when you’re competing for a customer’s attention.

You need to learn how to wear multiple different hats and digest different points of information quickly, and react to it in line with your brand’s proposition. That’s easy to do all at once, right?

Managing Expectations

Trying to strike a balance between volume and efficiency targets (CPA/ROAS). I don’t think there will ever be a perfect solution to this, aiming for one often leads to less of the other to some degree. For example, if you’re trying to get the most possible conversions in a given month, you will need to be more aggressive, entering more auctions, testing out more campaign formats which brings more risk and then an associated higher CPA or lower ROAS.

We work with the client before creating any campaigns to understand their business goals and where they sit on the scale of volume vs efficiency, usually leaning slightly more one way or another. This gives a good starting point and allows us to better set expectations, then we can course correct as we go, if we need to, once we have the live performance data. 

Post-click

There are issues post-click. For example, if the landing page you’re sending people to doesn’t convert on any channel or has had very little success, all of the ‘tips and tricks’ or ‘hacks’ in Google Ads won’t make a difference. This can happen if the product or service is new to the market, if it has been successful elsewhere but you’re now trying to promote in a new territory and then general UX issues causing too much friction. We usually look at this before we launch activity to try to mitigate risk but failing that, if your campaign side metrics look good such as CTR and you’re just not getting any conversions, you can look at basic metrics in Google Ads such as the landing page experience component of quality score, use auction insights to review your competitors websites for comparison or if you’re an ecommerce business use a tool such as Grow My Store to get more specific insight.

Conversion tracking

If the right data isn’t getting back into your account this is a huge problem for Smart Bidding alongside general ROI calculations and reporting. Often people are tracking multiple conversion actions with different values, for example, click to call, newsletter sign ups and sales which is fine but if they’re all being included for Smart Bidding, as a PPC specialist, you might be being measured on sales but optimising for phone call clicks. Another common situation I see a lot is a conversion double firing which leads to over reporting on ROAS and general performance.

All of my team are trained in Tag Manager and analytics and my advice would be to ensure that you and your team are too. This means that we can ensure we’re feeding in the right signals before our campaigns go live and if something drops out which can often happen when changes are made to a website, we’re able to pick up on it and diagnose the issue. The last thing you want is to be ripping your hair out optimising your campaign, wondering why it isn’t working to find out that the tags weren’t firing correctly!

Machine learning will continue to become an ever more prominent part of the ad platforms which will likely mean more simplification campaign formats and less “hands-on” PPC stuff to do. This is great news for businesses as it means easier ways to get their products and services in front of customers online. For us, it means we need to continue to evolve with the platforms but also learn to add value in other areas such as understanding value proposition, conversion rate optimisation and also being able to use all of the tools at our disposal. We don’t talk about “PPC” so much anymore, a multi-channel paid media approach is usually where we see the best performance for most clients, diversifying away from just search is a huge value add for yourself in your career and also your client’s business growth.

For me, attribution is probably the biggest; most conversion journeys contain multiple touch points across different campaigns, ads and channels.

This makes being able to show a direct return very challenging a lot of time, you can’t necessarily discredit any given input but it’s also hard to then show that they all contributed from the perspective of what went in, what came out. For example, all the ad platforms will take credit for the sale or the lead if they contributed, however, in the back end of the site there isn’t the same total volume.

If we’re talking Google Ads then always data-driven where possible, if the account isn’t eligible then one of the non first/last click models such as time decay, linear or position based.

Data driven is the most accurate way to attribute conversions in the account. Outside of this one of the non first/last click models at least get you some of the way. First and last click models ignore a huge amount of data which can lead to ineffective or incorrect optimisations and certainly poorer performance when using Smart Bidding.

Conversion journeys are complex, there are often multiple ad clicks, for example, that lead to a single conversion, using first/last click means all of that journey data potentially never makes it into your bidding strategies or your considerations for other optimisations and changes.

Micro-managing campaigns

Let it go through the learning phase and make changes only when you have a statistically significant amount of data, let it learn again and then review. Consider time lag in reporting, reporting on short time frames such as weekly or daily often isn’t giving you an accurate picture of the full value your spend has returned.

Expectations being mismatched 

ML and automation in PPC is awesome! It won’t work under the wrong conditions though.

Max conversions will try to spend all of your budget every day so if you think you can set it and forget it… you’re wrong. Same for efficiency strategies, it would be amazing if we could put in a target CPA of £1 and expect it to work but if that’s not inline with the auction and your actual performance it’s unlikely to work. Set your targets in line with your current performance then look to reduce them gradually over time.

Being too “in the weeds”

It’s easy to get swept up in someone else’s cool targeting ideas or ‘best practice’ and not focus on the bigger picture.

Those ideas may have worked for that person and they may work for you but it’s no replacement for core strategy and an understanding of the business you’re promoting. See testing like this as an addition to your ‘always-on’ activity rather than a replacement or subtraction.

I think the biggest mistake I’ve made in the past has been not knowing when to accept that something hasn’t worked. Not everything you’re going to try, is going to be a success. This led me into a place where I wasn’t taking my own advice and wasn’t looking at the bigger picture so I started micro-managing the campaign trying to change every detail every few days to get it to work which just added to the problem. A negative test can be as useful as a roaring success of a test, knowing when to say your idea hasn’t worked and to learn from it and move on is the key differentiator to success, in my opinion. Ok so it didn’t work, but why? What could be different and tried again? What happened the 2nd time?

I think losing that fear of failure type mindset and being more scientific and objective will help you to stay focussed on success.

I love what I do so I’m not sure I have one to be honest. Maybe if Google rolled back to only having manual bidding…… joking just trying to wind up the PPC community ;)

I’m not sure! What I am sure of though is things will change, we will adapt and things will continue in some form or another!

We saw a lot of people pull out of auctions when the first lockdown was enforced here in the UK due to uncertainty. Consumer demand didn’t actually drop in many of those verticals though and it actually increased in many instances. For those who were able to keep advertising, they reaped the rewards! Being data led and business goal focussed is essential to make reactive decisions more than ever.

I would love to read Frederick Vallaeys answers to these questions.

The key to success for a PPC specialist is “to do more than PPC” and we totally agree on that.

We thank Byron for his time, valuable thoughts underlining major PPC problems and examples how to solve them.

We hope this interview was useful and you enjoyed this read.