Google’s new PPC features are loaded with AI

At Google’s San Francisco marketing conference, held in May, the internet super-corporation made some startling claims about the future of its AdWords tool. Of huge interest to anyone who engages in pay per click campaigns in the UK, the event – named Google Marketing Next – saw a number of innovations added to Google’s current offering. Aimed at larger businesses in the main, their presentation of key new features was certainly designed to be eye-catching, but what of the details which will be of more interest to smaller organisations? Read on to discover the future of pay per click marketing and how artificial intelligence (AI) and machine learning will play an ever greater role.

New Customisation Features Help You Hone In

In the near future, Google will allow e-marketing professionals to perform a query that will let them see who, or what sort of surfer, is accessing their website via a pay per click campaign in AdWords. This means that for the first time it will be possible to tell whether a particular search term is being clicked through on from a genuine customer or perhaps someone simply doing a bit of research. Google’s Director of Audience Products, Bhanu Narasimhan, said that this function will allow for conversion rates for in-market audiences that are improved by ten per cent, on average.

Any in-market search feature is likely to improve the ability to attract specific types of consumer. However, this is not all Google have announced. At the present time, there are just under 500 types of in-market audiences that can be displayed within AdWords. In a further announcement made by Karen Yao, the organisation’s Group Product Manager for Ad Platforms, fully customisable in-market audience will now be possible, too. For example, by adding keywords that might have been used by someone looking for a product or service you sell, it will be possible to customise an in-market audience specifically for them. Thanks to the big data and machine learning offering of Google, this should allow for a much greater targeting of customers, new and old.

Target Your Clients’ Life Events

In the past, life event targeting relied on fairly simple AI, but Google’s Tensorflow processing system has revolutionised it. Getting to grips with would-be consumers who are going through a targetable life event means being able to gain a distinct advantage over competitors.

Whether you market to recent graduates, to people who have just bought a home, or to newlyweds, advertising to people in a more structured and, frankly, intelligent way is now available due to Google’s investment in this fast-moving technology.

Use Attribution Modelling Features to Track Customer Interactions

Google Attribution is already available in AdWords, DoubleClick and Analytics, so what needed to be improved? Basically, the issue with many attribution models is that they are somewhat limited when it comes to reproducing real-world behaviours. People just don’t ‘behave’ online in the way that we assume they might. Nevertheless, thanks to Google’s updated store visit data, better store sales data and a consolidation of data that is easier to read, marketing professionals won’t need to try to use the existing tools in the same way anymore. For example, constantly tweaking models for one-off or seasonal behaviours will be done away with. Due to Google’s machine learning, data-driven models will now cleverly weigh up how each click point in a sales and marketing process contributes to the overall outcome.

In Summary

There are four key points to take away from Google Marketing Next. They are:

1. The growth of artificial intelligence in the way Google will run services like AdWords
2. The increased accent on life event targeting.
3. The ability to customise in-market audiences.
4. The fact that automation is likely to play an ever greater role in e-marketing this year and next.

Therefore, it is going to take even more specialisation and expertise to successfully squeeze the best out of e-marketing campaigns in the future. It looks like another learning curve for agencies and another reason why you shouldn’t attempt to use all these tools yourself!

The bottom line – why nothing else (really, truly) matters

Lies, damn lies and statistics is a well known expression and whilst we don’t strictly adhere to the hostility of this line levelled at what is just an inert mathematical measure, it’s instructive that figures can get a bad rep if used in the wrong way.

We’d like to outline why you need to sweep away the negative thoughts and embrace measurements and assessment criteria as they are the lifeblood of any ambitious or growing company.

Picture this: your financial manager comes into see you one day saying we need to invest more money into a new marketing channel and when you ask:

‘how much?’

They reply ‘I don’t know’

I don’t think it would be likely you’d grant their wishes. Marketing no longer needs to be conducted in a ‘hit it and hope’ fashion.

So why, fundamentally, does the bottom line matter so much? To understand this, lets take a look at SMART objectives, KPIs and measuring performance.

SMART-en up your thinking

Smart goal setting concept

Any growing business needs to be closely and carefully monitored and managed.

As your business grows you need to understand where you are today and just as importantly, it allows you to set the scene for target setting to influence your strategy for future growth.

As most marketing managers will be able to tell you, the best types of objectives are the ones which are SMART (Specific, Measurable, Achievable, Realistic and Timely).

A non-SMART objective is no objective as the saying goes.

KPIs are the business

Key Performance Indicators or KPIs, are the specific performance measures for your business that you choose as your benchmark. They are your key business drivers and by definition, the parts of the business you should choose to focus on. You should be able to determine what your KPIs are. Put simply, the performance of these has the greatest impact on your business success.

Which leads neatly onto the digital world and the need to be able to measure the commercial benefit for any strategic or tactical digital activity.

There’s an apocryphal story of an advertising man saying:

‘I know that 50% of my advertising is working (and by extension, 50% isn’t!), the only problem is, I don’t know which 50%!’

That is no longer the case, with digital we have an opportunity to determine with both accuracy and confidence exactly what works and what doesn’t. And in the digital world, you really should be aiming to cut out the stuff that’s proven not to work, and maximise the stuff that is.

It sounds desperately simple, doesn’t it? Which makes it all the more surprising that many digital agencies seem to be focus more on the visible symptoms rather than the underlying disease, metaphorically speaking. By this we mean being more concerned with marketing activity metrics rather than what these metrics mean to the commercial interests of the company. Which seems very odd to us.

Measures that allow you to perform

When it comes to results and metrics and smart objectives, it pays to consider a company which first and foremost measures the bottom line for our clients (and is judged accordingly). Which is also why one of our challenges, to you, is to ‘do more’ because by doing more, and measuring more, you are more likely to achieve a successful result.

Which is precisely why most of our illustrations and case studies measure the commercial difference our digital actions made to our clients, rather than anything else.

‘Result’, as you (and we) might say!

Big data and the role of decision software

Even supercharged data has a job to do, and the job is always about a decision – so says a recent MIT article ¹

And who are we to argue. So whilst the value of decision software’s promise has never been in doubt, the application of such was a different matter entirely.

So what’s changed? For one thing, the sheer scale and amount of data available electronically has given extra impetus to an automated solution. Additionally, the reduction in complexity of using the necessary expert systems and the development of open sourced software means that for the first time, real automated decision making is now within the grasp of many businesses.

OneDecision.io_logo_portrait

Leading the charge is Omny Link, Ergo Digital’s technology arm, and with the development of One Decision, a collaboration between Omny Link and industry heavyweights Alfresco and Signavio, this early promise is now being realised. So much so, that a big 5 consultancy firm is working with us on a national European project.

Which industries are suitable?

So what sort of applications and industries lend themselves to decision management software? The most appropriate situations, unsurprisingly, are the ones where decisions need to be made frequently and rapidly. The knowledge and decision criteria have to be structured to a high degree and include a thorough understanding of the factors to be taken into account. If you can codify the decision rules and have access to high quality data, then away you go!

The obvious industries ripe for implementation include insurance (particularly underwriting) law, banking and travel. But the applications for this technology are only now being seriously realised. For example, within the marketing roles of SMEs, provided one is able to define the rules engine (a set of business rules that use conditional statements to address) the organisation in question would be able to apply automated decisions to their day-to-day working thereby enabling them to do more with less, and freeing up their managers to concentrate on their more important activities.

Go further

In the next blog, we’ll discuss how coupling decision software with workflow applications via the cloud and API connectivity is enabling dynamic companies to manage their workflows more effectively.

¹ MIT technology Review January 2014

We’re the GPS for your marketing journey,
guiding you to success.

Map Your Success