It probably seems weird to publish a piece titled like this after such a successful Beacons recently, but I’ve made a lot of mistakes in the last four years. Some of them were unavoidable (but taught me a lesson), some should have been avoided. Almost without exception, these happen because of rushing, or lack of process. Hopefully, this article will help you to avoid making the same mistakes – and remind me to repeat them a little less often, too.
I’m very driven by achievement. In digital, you can see the immediate impact of the work you’re doing. Everything can be optimised, which is like crack for people who are achievement-driven.
This sometimes reminds me of the Paperclip maximiser thought experiment from philosopher Nick Bostrom, where artificial intelligence is tasked with maximising the number of paperclips in its collection. It quickly spirals out of control and it ends up consuming humanity in its quest to produce paper clips.
We’re not heading into a philosophical essay relating to AI, humanity and paperclips, but it’s interesting to think about in relationship to our industry and our specialist services – the analogy of the siloed and one-dimensional thinking still resonates.
The danger here is to over-optimise towards goals that are too specific or relying too much on narrow fields of expertise. The tools we use are very good at optimising the numbers in their specific domain, but to ultimately get good results you must have a more holistic view and work across disciplines.
Here’s an example: (admitting that may be my next big mistake) I’m not really sold on bidding on your own brand terms in SEM. In most cases, there is already an un-paid organic listing at the top and there’s no need to put a paid ad there, too. (In search terminology this is called ‘cannibalization’.) If your competitors are bidding, you ‘need’ to be there, but it’s a zero-sum game. (There are also instances where bidding on competitors brands work, but again, most of the time that doesn’t make sense either.)
Most SEM specialists will completely max out the impression share on brand campaigns (and mostly with very good intentions). I did this too, because when you do, your results via SEM skyrocket because most people enter and convert via brand terms. It’s best practice in every SEM playbook to do this. The thing is, those aren’t incremental conversations. The brand was already getting those conversions via the organic channel and your ‘skyrocketing’ SEM campaign just meant you paid money for something that the brand previously had for free.
We have a lot more testing to do on this relationship, but in every instance we’ve tested to date, we’ve observed not just a lack of incrementally but in many cases, a net detriment to online sales where there is a high SEM impression share on brand terms. I don’t think the answer here is as binary as do it or don’t – there are many other permutations of tests to do. The point is if you’re in a specialist role you can be actively motivated to make your own results look good – often at the expense of another more important metric.
Identifying these issues isn’t especially hard in search, but it’s incredibly difficult in other media. Fixing them is more difficult because that requires both challenging objectives and having high-trust relationships with colleagues and clients.
Setting the wrong goals/not challenging objectives
Challenging objectives is probably the trickiest part of my role. Most clients have KPIs per channel and asking them to not worry about siloed results and focus on net revenue – or stop looking at a CPA and measure value of sales – is a really big ask. I really struggle with the decision to not change the KPIs when I know they’re just vanity metrics, but sometimes that’s just how it is. The important thing is to try. When it comes to challenging objectives, there are many wrong ways to do it (and I’ve discovered most of them it would appear) and the biggest learning really is that you need to offer suggestions of what the objectives should be (not just challenge the ones given) and if you’re going to do it you need to do so proactively.
When this brand SEM issue occurred and I realised what was happening, I was solely focused on the bigger picture – not the given objectives. So, when we lowered our brand share our total online sales grew – and predictably, SEM results ‘tanked’. I didn’t have a strong enough relationship with my clients, or their trust, to communicate that what we were doing was more effective (see section ‘failing at building relationships’) and consequently caused a fair bit of negative sentiment towards myself and the team.
At the time I thought this was very unfair but in retrospect, it wasn’t. Total online sales weren’t our objective and whether it should have been or not, we needed to have challenged that proactively and taken the client on the journey with us.
“Is it smart or is it just complicated?”
It’s something I have written on a post-it note under my screen and historically, I’ve gone against my own advice on many occasions. There are three main reasons this happens:
- Pressure of ‘innovation’.
- Pressure of ‘test-and-learn’. We should always look to continuously improve. However, we need to do this in a measured, sensible way, otherwise all we test is noise. This causes unnecessary complex setups, deters from what’s important, and is often counter-intuitive to how the tools work. It happens when we’re not proactive in our approach to continuous improvement.
- Sprinting before we can walk. We’re not short on sharing successes with each other in media. We see others’ final product and with the best intentions we want to race to be there too while forgetting about how much time, process and learnings they’ve had to get there. This often leads to massive complications and bad process which causes mistakes.
To fix this requires strong relationships above all else.
Building relationships with colleagues and clients
I used to be so shy I wasn’t brought to client meetings. I wasn’t ‘client-facing’. I was good at the ‘doing’ of my job but I couldn’t communicate my ideas alone, and I relied on leveraging my colleagues’ trust with clients to sell in ideas. The more I did this, the worse it got – much like a self-fulfilling prophecy. I was so interested in the work, all that ‘relationship-stuff’ was secondary and I thought it was pointless. Suddenly, situations arise where you need to do things yourself (see above three sections).
I’m still really shy, but I’ve realised that’s okay. When I realised that lots of the people around me were actually introverts, too, (even very senior people) things became a lot easier. I used to think everyone around me was born with charisma and I simply wasn’t. It’s definitely a skill which some people learn faster, but everyone can learn it. Demonstrating vulnerability is really powerful. Let people get to know you and make sure every person you work with knows you both like them as people and respect them as colleagues/clients.
It’s impossible to overstate how important it is to build relationships with those around you.
Asking for help
A lot of people entering the industry will have the same attitude I had when I started – that being ‘bright’ and hardworking is enough. (It’s possible I was neither bright, nor hardworking enough to have made this approach work.) Either way, I found you’ll inevitably need some help somewhere along the way. I needed to learn how to ask for help. A lot of my mistakes were just because I didn’t ask for help. There’s not necessarily a good or bad way to ask; you just need to ask.
Something I learnt from this is that there’s also more you get out of asking for help than solving an immediate problem on hand.
Most things you can technically ‘solve’ yourself by reading, chatting to DV360 help desk for hours, or my favourite past time: writing out the problem in a fake email to Ivan (our head of technology). Usually, by having explained it in as logical a way as possible to someone who doesn’t have the context, you’ll have solved the problem in the process.
Asking someone else for help gives you much more than a fix. It shows you another way of solving something you couldn’t ‘solve’ yourself.
Asking for other perspectives
If you’re not confident enough in your own perspective, it can be daunting to ask others for theirs, without feeling concerned that it’ll somehow reflect badly on you for not having your own perspective.
But only talking to your closest colleagues and not widening the perspectives you hear isn’t helpful for you. The very best thing about working in an industry like this is that you’re surrounded by a lot of perspectives and a lot of talent so make the most of that.
It would be fair to say I’m mildly obsessed by the packaging of information and it would also be fair to say that people get a bit irritated with me about it.
The way your documents look or how you present yourself has a very direct relationship to how good most people think the content of it is. A badly presented document or a poor presentation – regardless of how good the ideas and content are – is far less likely to be read or actioned. If you aren’t taking a lot of care in the presentation of your work, you’re doing yourself a massive disservice.
I’ve never searched for cheese. I’m not subscribed to any cheese emails and aside from some cooking shows, never watched cheese-related content on YouTube (or any other video platform). Yet, I buy a lot of cheese. Most of the things I buy or use daily I’ve never searched for. Until one of our major supermarkets decides to sell their user data – I probably will never be ‘in-market’ or have an ‘affinity’ for cheese products. We’re not far away from this kind of data availability, but it’s not here right now.
If you’re running a retail campaign, the temptation to rely on these audiences alone rather than speaking to a mass audience is understandable. I’ve been there: “Why go mass when we can market directly to people who are already shopping?” and the simple answer is because it isn’t as effective as doing both. Even if your client sells stuff which is searched for a lot or is an online business, most of the time you should still do both targeted and mass communications.
A lot of the tech isn’t so shiny under the hood
I read this in a piece from Tom Goodwin recently: “Most systems and software are an amalgamation and aggregation of patches and quick fixes, hacks built on temporary workarounds augmented by kludges. Things that work on a good day, just about, but nobody quite knows why.”
Lots of the tech used in advertising seems pretty state-of-the-art. And some of it is. Some of it is also not so great. Or at least, not what it is made out to be.
I’m not going to call any out, but many mistakes have been made by putting too much faith into these tools ‘just working’. As agencies, we must ‘make it work’ and this is where the “aggregation of patches and quick fixes, hacks built on temporary workarounds augmented by kludges” comes in. What was a smart fix to begin with ends up with band-aids on top of band-aids until the whole architecture is too complicated for anyone to sensibly grasp.
This is impossible to ‘fix’, but the point is that we should focus on more than just ‘making it work’. Our value as agencies and individually as performance specialists needs to evolve further. We need to get better at informing our clients about the limitations of tools/products and we need to get better at not jumping to use every new tech, every new beta. Don’t let innovation be a distraction from good work.
I was very grateful to have won the 2019 Beacons Rising Star award. For every success which helped me on the way, there were many more mistakes, and the successes were generally only possible because of the rest of the team here. I’m incredibly grateful to everyone who’s helped me solve problems, shown me a new way of looking at things, been a cheerleader and, occasionally, given me a bit of a ‘kick’ when needed.
For everyone on the way in the industry, hopefully this helps to show how many mistakes I’ve made, and help you avoid some. Ultimately, I hope it’s illustrated the most important learning: to find a workplace that supports the making of and learning from mistakes.