Look Out for These 5 PPC Trends in 2018
As we enter a new year I wanted to throw together some of the reasons I’m looking forward to PPC in 2018. Some of the big changes from 2017 (though there have been many which you can read more about here) affected all aspects of daily life in the industry from reporting to optimisation and even campaign creation/strategy.
- New UI
- Expanded Text Ads permanent
- Online-Offline Tracking
- Google Attribution
- Google Exact Match Changes
Although the summary above was just the tip of the iceberg in 2017, there’s no sign that PPC in 2018 will slow up and the 5 big trends below are likely to gain momentum and affect marketing across the spectrum.
Pay-Per-Click professionals jobs remain safe and growing in 2018, however, machines and, machine learning will continue to change the role and key functions of PPC and PPC professionals over the next few years. Jeff Baum points out in a recent article how a range of new features make direct use of machine learning in a way that simply can’t be done manually by even the most skilled professionals. Some of the most recent of these are:
- Smart Bidding –
- Smart Display –
- In-Market Audiences –
The whole direction of Google & Facebook are geared towards building AI and Machine Learning into every aspect of their products and, their own advertising platforms will likely be full of tests and firsts. The simpler changes (by which I mean the ones that non PPC professionals are likely not to notice) such as Google changing the meaning of close variants are completely fuelled and made possible by machine learning advances.
Google also recently trimmed the ad-rotation options available in AdWords to make better use of what “it” knows and further move advertisers to relinquish control of the big number crunching required to run a solid a/b ad-copy test.
How To Be Ready
Despite these steps forward in machine-learning, I still believe my prediction for 2016 to be good advice. The goal here has to be to help the machine learning do the best job it can, using our skills to allow a machine to go above and beyond what would be possible by one alone.
Let’s take the ad-optimisation use of ML as an example, at least in 2018 AI doesn’t have the skills to really impress in self-writing ad-copy, there are too many variables and copywriting is still a way away from automation.
So when I set my campaign settings to “optimise” ad rotation, what I’m really saying is, “can you please use your big robot brain to help me pick the best ad for the time/place/search based on factors I’m too busy to really look at.”
But really, all we’re teaching the machine here is that one of my ads the worse, instead of me writing 2 or 3 variations (which PPC analysts have been looking at since the ancient days of 2012 and before) I need to be feeding that big computer brain with way more information. Writing 5 or 10 variations and throwing serious traffic at it, in the knowledge that the more I give it the better it gets in comparison to my puny mind needing several macros just to make sense of one split test.
The same applies to keywords, on most accounts, the relative gains of splitting out every keyword into a SKAG are mostly overrated, with little data going to many keywords making it more difficult to draw conclusions. Within reason, a better use of time would be writing more ads, keeping the sentiment/theme appropriate, so the right ad shows and then, letting the much smarter (than me at least) algorithms do their job of optimising the performance across the keywords and portfolios.
Closely linked and, enabled by the advances above in Machine Learning, we’re really going to see the end of last-click attribution as a model and the continued growth of attribution as a service or product.
Last-click was easy, it got us from A to B. However, if you’re hung up on this in 2018 it’s basically certain you’ll be throttling your growth by under valuing earlier touchpoints.
Google Launched Attribution, as well as re-jigging the whole analytics suite of products. Additional insights are available in cross-device conversions which, are now included as default. Plus, Data driven attribution (DDA) rolling out in AdWords means larger advertisers can leverage the power of ML (see above) to make better attribution modelling easier than ever before.
Facebook has upped their game in including importing offline conversion data and both Facebook and Google are now also providing in-store or offline attribution in some applicable cases.
How To Be Ready
If you’re using GA go ahead and take a look at your model comparisons. It’s not really about choosing one attribution model and sticking with it come hell or high water. Think of the situation as more about clearly understanding how different attribution models affect the answers to your most common questions.
Voice & Conversation
Queries and interactions with ads will continue to change as voice search continues to grow. This is more than just a gimic and no longer is it about that one guy you know who uses Siri for everything.
In 2018 it is predicted 30% of all device interaction will be voice-based, this adoption will be even greater among certain demographics.
Worried it’s not monetized directly in AdWords yet? Remember that Google Shopping, then called Google Product Listings were free too .
Earlier this year a lot was made about Google’s apparent lack of plan to generate revenue in a world which isn’t looking at their search results page. However, YouTube and Gmail weren’t initially monetized either and the recent tests for ads on Google Home show that voice ads are well within reach of Google, whilst Facebook are stretching their reach more with conversational chatbots with innovations like messenger ads we continue to see a shift to different types of queries and different ways for customers to discover products and services online.
How To Be Ready
Make better use of BMM, there’s simply no way to stay on top of the search query variations that are being generated for even fairly simple root keyword combinations so use the advantages you have where possible with dynamic search ads, dynamic keyword insertion or ad customiser to try to scale up your coverage to account for the many keyword variations growing in 2018. Think question or recommendation based with your ad copy and keywords as a starting point and analyse your search query reports (SQR’s) from there.
Nearly three-fourths (74%) of online consumers get frustrated with websites when content (e.g., offers, ads, promotions) appears that has nothing to do with their interests, unsurprising, though perhaps easily ignored.
Does this ad look good to you?
But what we’ll need to be thinking in 2018 is. “Does the site provide what person X will need at when they search at this exact time and place, and what about the next time they look for it? Customers expect more personalised recommendations across the board and, their expectation from PPC marketing will be no different.
Growing ID based targeting like customer match in Adwords or custom audiences in Facebook is also being augmented by other services such as Live-Ramp in addition, the expansion of more targeted demographic, affinity and in-market segments becoming available to search markets such as Google “life events”
How To Be Ready
Leverage “micro-moments” with the features available to you, and make use of your audience data by adding it as an audience list at campaign level as soon as possible, doing this on the “observe” setting means you can start segmenting in your reporting before doing any optimisation.
- Make It Routine
- Tailor Your Copy
- Expand Your Campaigns
Look at demographic and audience based bid optimisations as you would with device, or geo optimisations and make it part of your routine. Although machine learning might be better at adjusting your bids, understanding “why” a younger demographic don’t respond so well to your call-to-action like “call now” gives you the advantage. Try pushing “chat with us live online” as an alternative.
Expand your existing campaigns by creating broader, top-of-funnel campaigns targeted only at previous visitors or those with an affinity/in-market affiliation.
2018 will be a first in that, for the first time in a while (at least for small to medium size businesses) Google will be given a run for their money in terms of ad-spend. Facebook & Amazon both increasing their ad formats. Facebook have introduced lead ads, store visits, dynamic product ads (DPA) and more to tempt more advertisers and go well beyond a place for branding with a huge amount of direct response campaigns also flocking to the platform. Perhaps more quietly, Amazon has also been aggressively growing their advertising business with 51% YoY growth although that’s infinitely smaller than the two behemoths of digital (Google & Facebook) it’s on par with LinkedIn and, since I haven’t seen any steps forward for Bing/Microsoft in 2017 in ad performance or vision I’d put Amazon above Bing/LinkedIn as the closest competition at the moment.
Fuelled by and, in response to the above, I expect to see ad-spend increasingly shared amongst non-google (and Google non-search) campaigns.
Lead ads have been a huge success for Facebook & LinkedIn with rave reviews, the synergy between Facebook Store visits Campaigns & Google rolling out Store Visits metrics to AdWords campaigns demonstrates how keeping pace with metrics & formats is going to be a continuing theme. I won’t be surprised to see Gmail Ads be presented with some new formats (dynamic Gmail Ads, Gmail Lead Forms) and a potential re-introduction of the form fill extension from back in 2013.
How To Be Ready
Don’t focus so much on what you do as a business, focus more on what your customers do – the format is the last thing to worry about when working out a campaign, first assess where your customers are, where you’d have to be in their journey to influence that best and then decide if you’re going to do it using a facebook messenger ad or LinkedIn sponsored content. Ad-spend will continue to be broken up by channels and by more niche campaign types, so better to think in terms of objectives than features, as you never know what might be available in a few months time.
I think great useage of feeds in PPC were more than possible in 2017 with the right approach and tools but, will continue to be a big theme of 2018 where the right approach to feed based campaigns will allow the right types of advertiser to scale up their efforts massively.
PPC Prediction 2017
Dan Pillay, Search Products & Innovation Manager
“Human Teach, Machine Learn” – Machine learning algorithms have made their way into both the trendy world of start-ups, as well as the pitch decks of many agencies.
I see the importance of automation of machine learning growing throughout this year in both PPC and many other sectors. HOWEVER, I believe the real breakthroughs in this area will be in the software being provided to allow humans/account managers/clients better ways of assessing and defining what success looks like to them and their business so that automation and machines can learn, optimise and iterate to bring positive results.