I've spoken a couple of times in the last year or so on the future of our industry and what it may mean for us. I wanted to write up my thoughts on this, share a few trends and give some pointers on what I think we should be doing as marketers to take advantage of these trends.
Obviously I'd like you to read all of this post, but to give you a tldr - our skill sets need to adapt and as marketers, we need to do our best to sit at the intersection of these three areas:
Before looking at the future and exploring this skill set, it's important to look at what's happened in the past because it can give us some clear pointers on what could be coming next.
Things used to be easy
When I started working online back in 2004, things were pretty straight forward when it came to ranking well in Google. They basically needed to worry about three things:
- A few hundred words of content on a page - it didn't have to be great quality content, just a few hundred words was fine.
- Your keywords scattered throughout the content - you would shove in the keywords you wanted to target for that particular page and not worry about the content reading well. You'd often just need to create a page for each keyword you wanted to target, even if keywords were quite similar to each other.
- External links - you needed to point anything from a few hundred to a few hundred thousand links to your website, again, they didn't need to be of great quality, you could pretty much build any links you wanted and they'd work.
It was pretty easy and what made things even easier was the fact that universal search wasn't as prominent back then. Image search, videos, shopping etc were still in their infancy or not introduced at all yet. So results were quite often ten blue links.
This meant that getting a website ranking and generating traffic was quite straightforward, hence the reason I had a number of affiliate websites making me a bit of money to pay for drinks while I was meant to be studying at University!
Things started to change
A few years later, I'd found my way into a job doing SEO full-time and eventually found my way to London to work for Distilled. It was around 2010 and the first major signal of things changing happened with the caffeine update from Google. This wasn't a penalty update or anything like that, it was a fundamental change to the Google index and how it worked. It allowed them to index the web at a much faster rate than before, leading to fresher results.
What we didn't realise at the time was that this update was a clue as to what was coming next. Not only did this rebuild of the Google index allow them to process crawling much quicker, but it also allowed them to process web spam and identify low quality pages at scale.
Panda in 2011
In early 2011, Google released the Panda algorithm into their index which sought to identify low quality pages at scale. It was based on a series of questions that Google had run machine learning against and been able to answer for every page they processed. These questions covered a range of things, but the scary thing was that it was almost as if Google was now able to predict a human reaction to something such as not trusting a website.
Not only did a large number of websites see their traffic from Google fall massively, but we saw a change in attitude from Google. Up until this point, they were always quite cautious with algorithm updates and would often let spam go unpunished because to punish website en masse, would possible hurt websites that hadn't done anything wrong. Panda signalled a change in this attitude because there was definitely collateral damage which many websites being hit who shouldn't have been. For the first time, they showed that they were prepared to take mass action against low quality websites and accepted that they may make a mistake along the way.
Overall though, the Panda algorithm seemed to work well and was definitely refined over time. The following is a graph from a website that was hit by a Panda update in 2014 and shows their organic search traffic:
Now this graph shows their revenue from organic search over the same time period:
As you can see, revenue from organic traffic doesn't dip and actually increases slightly.
Why?
One theory is that before the Panda update, this client were actually getting traffic that they shouldn't have really been getting. This was because they were a very big, very authoritative site and therefore, lots of pages ranked relatively easily for keywords that weren't really that relevant. Therefore, the traffic they lost wasn't great quality in the first place, hence the lack of a drop in revenue.
Penguin 2012
After Panda, the next logical target for Google was links. It was still pretty easy to manipulate Google using links with techniques such as article marketing, directories, press releases and blog commenting being very effective. The thing that made things easier was the fact that the worst that was likely to happen if it went wrong was that it just would work. You'd have to be really, really aggressive to pick up a penalty.
Then Penguin came along and changed all of that.
While Penguin targeted a range of spammy techniques, including over-optimisation of on-site elements, in the weeks that followed it was clear that Google had come down hard on link building techniques such as those mentioned above that were scalable.
All of a sudden, any link building techniques that were scalable were a lot more risky and more likely to lead to a penalty than before. They could still work (and some still do today) but the risk factor meant that SEOs started to steer clear of these techniques. I actually believe it was this that led to the growth in content marketing services from SEO companies, but let's talk about that another day!
Hummingbird 2013
Following Panda and Penguin, which were focused very much on removing low quality results from Google, Hummingbird in 2013 was a very different update. Hummingbird was a move towards Google understanding what we actually wanted after searching for something, as opposed to just providing a set of results based on the keyword we searched for. They started to take into account what they thought our intention was and other implicit signals such as the device we're using or our location. My old colleague Tom Anthony talks about this in a lot more detail in this post over on Moz if you'd like to learn more about this.
Essentially, Hummingbird moved Google more towards conversational search because it could better understand what we were actually looking for and could provide the answer right there on the results page - not buried amongst the ten blue links below. Here is an example:
We can also see Google getting very smart in trying to give us the exact answer to a query. Here is another example:
What interests me about the result above is not the snippet of text being displayed, but the fact that the number "£6" is in bold. This is Google attempting to give me the exact answer rather than a chunk of text where the answer may lie.
What can we learn from all of this?
So the question now becomes, what does all of this mean for the future?
1 - Google don't want to be the middle man for free any more
The days of ten blue links on a search result are long gone. We're seeing Google take up more space on their results with ads and their own products and we can't expect them to slow down. We're seeing Google trying to keep users on their own properties for as long as possible, but it wasn't always this way.
"We want to get you out of Google and to the right place as fast as possible."
I'm pretty sure that even back in 2004, Larry Page knew what he meant by "the right place":
Because something else happened in 2004, Google floated on the NASDAQ and became a public company. This meant that all of a sudden, they were being held accountable to shareholders and had a duty to drive revenue and profit for those shareholders.
This is never going to slow down and for this reason, I think pretty much any online business needs to accept that paid advertising online is going to become less and less of an option.
2 - Google will keep getting better at spam detection and AI will play a part
We'll talk in more depth about artificial intelligence shortly and the role is may play in spam detection. The point here though is that Panda and Penguin allowed for spam detection on a huge scale and gave the web spam team a lot of power. Again, we can't expect Google to ever slow down on this front and with the development of AI, it stands to reason that the web spam team will be able to leverage this technology at some point. Much in the same way they were able to leverage the changes that the caffeine update made to allow them to deal with spam at scale.
I think that spam will always exist and work in Google. But the length of time it takes for Google to catch it will become shorter and shorter, forcing spammers into a consistent rinse and repeat pattern. I think this means that the top 1% of spammers will be fine and continue to make a lot of money online, while everyone else will find it too hard and frustrating to keep playing the game. For digital marketers, it makes taking risks with client websites even less likely.
3 - It's still all about search
Despite all the things we've seen Google do over the last few years, such as these cars, contact lenses and hot air balloons, it's still all about search. Search is their main driver of revenue and will be for a long time to come. Even the introduction of Google Now is still fundamentally about search - the input methods are different to typing words into a box but the result is the same - giving you information as you need it.
4 - If it's ever not about search, it's about us
If you're using a Google product and it's not search related, it's about you and your data. Think about Gmail, Google Docs, Android, Google Analytics, Nest and many others, they all gather data about us and the web. They can then use this data in a number of ways, one of the most obvious being to build audience profiles which can be used by paying advertisers.
I also know that one of the key things that Google Ventures look for in a company is the data they gather via their product or services and I saw this a year ago with a potential acquisition they were looking at in London.
5 - AI is getting bigger, whether we like it or not
It's an area that certainly divides opinion, but like it or love it, artificial intelligence is growing and is going to become a bigger part of our lives. As marketers, the development of AI is something we need to be aware of because it could fundamentally change how audiences interact with brands via online channels. We'll talk more about this shortly.
We've seen the first signs of growth and the usage of AI, the next few years will see faster movement because of Kurzweil's law of accelerated returns which you can learn about here:
So what about the future?
3 trends for the future
Now that we've looked at the past, what about the future?
There are three trends that I want to explore that I think all digital marketers need to be aware of. The truth is that none of us know what these trends mean 100%, but there is one thing I'm certain of which I'll talk about shortly.
Trend 1 - Artificial intelligence is growing
I touched upon this at the end of the previous section and I want to go into more detail.
We've seen in the last few years that large companies are investing heavily in AI with Google making a number of acquisitions to grow their talent pool and technology. But it's not just Google who are doing this, we've seen moves by Facebook and further work by IBM to invest heavily in this area.
In my eyes, AI is doing two key things that marketers need to be aware of:
- AI is filtering what we see online
- AI is predicting what we see online
Remember that as marketers, it's our job to connect brands to their audience. If AI is going to potentially get in the way of this, we need to take notice.
AI is filtering what we see online
So much information is thrown at us each time we look at our smart phones or laptops, we'd be overwhelmed if there wasn't any filters in place. It's gotten to a point where I'm sure that many of us don't even notice the various filters that are part of our favourite platforms. Here are a few examples.
Gmail and the priority inbox feature allows Google to guess which emails are most important to us:
It's not just the main inbox either, they also filter social network and promotional emails into a separate tab:
This means that even if you've chosen to subscribe to a newsletter for a particular company, their emails won't make it into your main inbox.
Our Facebook streams behave in a similar way. We have our primary stream where our friends updates are shown and the friends we interact with are shown the most, this is known by most people. The crucial point for brands is that even if a Facebook user likes your page and even if they interact with it a lot, there is still a chance that you'll be filtered out of their feed. This study by Ogilvy is often referenced and shows the gradual decline in organic reach for brand posts:
Remember that this was early 2014 and it's safe to assume that the reach is even lower now. What's interesting is when you map this data against the Facebook stock price during the same period, which is what Convince and Convert did:
Can we expect Facebook to give the organic reach back to brands when we see the above happening?
It's not just Facebook either, Twitter are experimenting with filtered feeds more and more which, I think, is important to their growth. Earlier this year they rolled out the "while you were away" feature which took a smart guess at the tweets we missed but want to see.
For brands who are active on Facebook and Twitter, this means that our marketing messages and content needs to get past these filters. It needs to break through and reach the level where an audience doesn't let these filters get in the way.
This also means that permission marketing is no longer good enough. Let's recap the points above:
- We can subscribe to a brand's email list, but Gmail will filter it out of our main inbox
- We can like a brand's page on Facebook and not see their content
- We can follow a brand on Twitter, but not see their content
We've given permission for a brand to market to us and we still may not see it!
AI is predicting what we see
Not only is AI filtering what we see, but it's also predicting what we see.
Not many digital marketers (including myself) knew who this guy was a few years ago:
I mentioned him earlier. Ray Kurzweil joined Google in 2012 as Director as Engineering. His brief was to focus on machine learning and language processing. When asked what he would be working on, Kurzweil said:
Discussing his project at Google…Kurzweil says he is creating a synthetic neocortex that can understand language just the way a human does.
Sound familiar? I'm not sure about you, but it reminds me an awful lot of what Hummingbird has done i.e. taking our search, understanding it, and giving us an answer. Much in the way speaking to another human being would:
The term coined for this is conversational search and you can read more about it here. I'd also recommend taking a look at the examples in this blog post, again from Tom Anthony.
I'm sure that Kurzweil had a big part to play in the rollout of Hummingbird, but his influence is not stopping there. In an interview with MIT Technology Review, he "talked confidently about making Google's current search technology obsolete." He didn't say this when he joined Google in 2012, he said this in 2014 - just 18 months ago. This is quite a bold claim, but who would bet against a guy with an amazing success rate at predicting the future? Although I should be fair and provide this challenge to his predictions.
Either way, Google are making huge strides forward in this area and it's not going to slow down.
Let's bring this all together.
In 2010, we saw Google make a big leap forward in how they indexed the web with caffeine.
In 2013, we saw them make a big leap forward in how they understood the web with Hummingbird.
In 2015, we're seeing them get even better at predicting what we want to see via Google Now.
So we've seen them move from Indexing > Understanding > Predicting.
Coming back to a point earlier regarding spam, AI will also make spam a lot harder. Remember that at it's heart, AI learns from us. The prediction models that Google are using takes it's inputs from our behaviour and when you think about the number of users Google has, there is a lot of behaviour to learn from. This is fundamentally how Google are able to do things like this:
They're effectively crowdsourcing this and seeing that these four possibilities are the ones that are most likely to be wanted be the searcher - based on millions of previous searches and behaviours. But the original input is from us.
This is why spam becomes harder. It's much harder to manipulate these inputs when they're powered by millions and millions of individuals. It's not impossible, after all, many of us can have our computers controlled without even knowing it. But again, only the top 1% of spammers will thrive and the rest will struggle to scale spam to an effective level.
You can also bet that the web spam team at Google will get access to AI technology when building their own systems to detect and prevent spam. Why wouldn't they?
Trend 2 - Mobile isn't really about mobile
There is no doubt that mobile is big, but I'm not convinced that we've been thinking about it in the right way. Let's look at a few numbers first to set the context.
We have seen that 2015 is the year that mobile search has overtaken desktop search for the first time:
In terms of what's driving this, it's not as simple as you may think. This change isn't through not having a choice, such as being away from a desktop, it's because we want to use mobile devices instead:
77% of mobile searches are in a location where people have a PC available to them
This was from a Google study which you can read in full here.
So this growth is our choice, it's not being forced upon us.
Facebook are also seeing big shifts in the way people use their platform. This graph from Ben Evans shows that in March last year, the number of users of use Facebook only on a mobile device overtook the number that only used it on desktop for the first time:
To be clear, this means that there are now a lot of people who use Facebook only on their mobiles - not on their desktop.
The easy conclusion is to say "desktop is dead" and that it's finally the year of the mobile. But this isn't the case at all. Let's look at some more data.
Mary Meeker's internet trends deck is a must read and has lots of cool data. But this graph is the one I want to draw your attention to:
The green line shows the growth in the amount of time we spend per day using our mobile devices to consume digital media. Pretty impressive growth right?
But look at the blue line which represents desktop. It hasn't declined and since 2008, it has actually increased slightly. This shows that desktop isn't dead, we're simply spending more and more time online and the majority of the growth is down to mobile devices.
One possible takeaway to this trend has been put forward by Rand at Moz with this tweet:
Mobile isn't killing desktop; it's killing all our free time. http://t.co/PbG3U1epzG pic.twitter.com/8hRWjXO9PW
— Rand Fishkin (@randfish) May 26, 2015
You can see that the graph tells a similar story to the earlier one from Mary Meeker.
I see what Rand means but I don't agree 100%. I don't think mobile is killing our free time, I think that we're using our free time differently. It used to be this:
Now, it's more like this:
The key here:
It's not about mobile, it's about context.
Sometimes, the context is that we're sitting in a bar, waiting for a friend, so we'll spend a few minutes checking Twitter, opening a few links and reading. Other times, the context will be sitting on the tube in London with no internet, so we'll watch a TV show we downloaded on Amazon Prime the night before. Finally there may be times when we're sitting at a coffee shop, need to do some work and the iPad keyboard isn't good enough, so we'll pull our laptop out and do some work.
Whatever the context, one thing is for sure - budgets are coming our way because of this behaviour.
Let's go back to Mary Meeker's deck and look at this chart:
There is currently a big deficit between time spent on mobile devices vs. the amount of advertising budget that goes into it. It's not a small deficit either - $25billion in the USA alone.
Who is in a position to manage this spend when it shifts? Our industry.
The shift in spend towards mobile is growing and despite claims that desktop is dead, it is also seeing growth in advertising spend but it is slowing down:
There is no reason to think that this growth will slow down and while desktop may slow and perhaps level out, it isn't going anywhere soon.
Google see the opportunity here too and we're seeing them put a huge focus on mobile, shifting engineering resource to mobile and even taking a mobile first approach to design with this quote from their Lead Designer, Jon Wiley:
Towards the end of last year we launched some pretty big design improvements for Search on mobile and tablet devices (mobile first! :-). Today we've carried over several of those changes to the desktop experience.
Dr Pete predicted this in 2014 and we're seeing it more and more. We can see that more and more of Google's earnings come via mobile advertising, along with Facebook's. So of course they're going to invest in this area and get a big piece of the pie.
It's not just about mobile advertising either. I think that for Google, a big play is trying to own the mobile experience and gather data from users as a result. They did acquire Android after all and start building their own phones.
Remember that even if you're not searching, using a Google product probably means that they're gathering data about you which you can then use in audience profiling and to show you future adverts.
To reiterate a point from earlier:
It's not about mobile, it's about context
With that in mind, let's look at the other big thing Google are doing to try and own the mobile experience and gather data about us - Google Now. Google have had this in mind since pretty much the beginning:
“My vision when we started Google 15 years ago was that eventually you wouldn't have to have a search query at all. You'd just have information come to you as you needed it.”
This is from a TED talk by Sergey Brin and sounds a lot like what Google Now attempts to do. But Google had a problem with Google now in that to get the most out of it, you had to be using an Android device and using Google's own apps such as Gmail, YouTube, Play etc. Therefore, their reach was pretty limited. Two things have happened in the last 12 months that opened up this reach.
Firstly, Google announced that they would start showing content from iOS apps in regular search results. They'd been doing this for Android for a while, but were leaving a big chunk of the market untouched by not doing the same for iOS. This means that iOS users are more likely to see results such as this:
This allows the user to click straight through into a deep page on the Open Table app. This provides a much richer experience for mobile users who don't use Android.
The second thing Google are doing, which I think is a very smart move is the development of Google Now into Google Now on Tap announced earlier this year. This allow them to interact with apps that are not owned by them. This brings down the previous restriction around Google Now only being useful for people that used Google's own apps a lot. Instead, they can do stuff like this:
Here, the user has the Spotify app open and then powers up a Google search, asking the question "who's the lead singer" and Google, now that they can "see" inside the app, recognise the context of what is being asked, recognise the band that are playing and returns the answer.
It's still not about mobile. With the example above, Google are still leading you back to search. But they're doing so from a place that isn't natively Google owned and can do it on a device which isn't their's either! Steven Levy sums it up quite nicely:
You know that phone you’re holding in your hand? It’s actually a search field.”
Trend 3 - Content discovery is mobile first
A lot of the inspiration for this section came from my old boss, Will Critchlow and I'd recommend checking out his various decks on the topic of mobile here.
In the previous section, we saw a lot of statistics that show the growth of mobile and we know that it's not slowing down. There is a big opportunity here for digital marketers which I want to discuss. But first, let's look at something important.
Despite the growth in mobile usage, it is still desktop that drives conversion:
Perhaps this is because retailers are still not providing a great experience on mobile devices, something that has some data behind it. I could also believe that some consumers are still not ready for mobile commerce and have trust issues. I wouldn't bet against the latter changing sooner rather than later with the introduction of Apple Pay and it's adoption in the UK.
Why mention this? Well it's another point that desktop is by no means dead and that's not what I'm saying in this post.
We traditionally think of content discovery like this:
Whereas actually, I think it's more like this:
The data backs this up too. These two graphs are from Aira clients and demonstrate the trend here.
The first graph shows two columns. The one on the left shows traffic to a client website, broken down by device. The one on the right shows the same website, but only content - i.e. we've removed the homepage, products, categories etc and only included blog content and a few pieces we did for them:
As you can see, the traffic from tablets and mobiles increases quite a bit. Here is another graph, from the same client:
Traffic from the mobile version of Facebook was much higher than the desktop version.
But it's not just about "viral" content when it comes to content discovery, there are all sort of content types that we need to be creating and making mobile friendly:
Research from Google suggests that mobile search is a key part of the decision making process, which shows how viral content isn't the only type that we, as marketers should worry about. The following statistics really hit home too:
All four of the activities above are things that we're trying to improve for our clients:
- Visit a retailer website
- Visit a store
- Make a purchase
- Call a business
If we can create content that potential customers can find via mobile search, they're more likely to lead to one of these outcomes for the businesses that we work with.
It's easy to think "cool, let's just make our websites responsive then" - but it's not just about that. Just knowing the latest technology isn't enough as this website nicely demonstrates.
It's about communicating the message in the right context. I'm not a designer, so I'd recommend taking a look at this deck from Vicke Cheung:
This gives some great tips for designing content with mobile in mind and making sure that the core message isn't lost in a sea of cool technologies.
So we've explored three trends:
- Artificial intelligene is growing
- Mobile isn't really about mobile (it's about context)
- Content discovery is mobile first
What do we do with all of this?
I love thinking about the future, but my pragmatic side doesn't like leaving a post without some actionable steps and takeaways, so here they are.
Our skill set needs to adapt
No one can predict what the trends above mean 100%, but I'm sure that our skill set needs to adapt to take advantage of them. We've seen this model for t-shaped marketers before:
I like this and I think it can work well. But I think it goes further:
Strategy - marketers who understand strategy and can see the industry trends and shifts and know how to react to them, will be at a big advantage.
Creativity - marketers need to be creative in order to break through the filters we saw earlier and to be the answer for customers who ask questions. Permission marketing is no longer enough and won't do this.
Technology - marketers who not only understand technology, but know how to use it to communicate to their audience at the right time, will be in a great position.
Marketers of the future need to develop skills around these areas and the ones who can position themselves as close as possible to the intersection of all three will excel.
Ok, but what do we actually need to do?
1 - Get really, really good at measurement
If we're going to take advantage of the budgets that are coming our way, we need to be able to measure the effectiveness of our work and show ROI for our clients. We need to understand the basics and understand cross-device measurement which is going to become more and more important.
2 - Start to work on customer lifetime value
As things become more competitive and a combination of organic and paid marketing becomes the default for many businesses, understanding lifetime value is more important than ever. Supporting organic marketing with paid advertising is going to be crucial and you can't just throw money against a wall without knowing how much you can afford to spend to acquire a customer.
3 - Build your own loyal audiences
Loyalty is more important than ever because of the filters that are being put in our way. Email marketing is by no means dead and this is often one of the most overlooked areas of marketing, but conversion rates from email marketing are as high as they have ever been, so it's a channel that can't be ignored.
4 - Combine paid content promotion with organic content promotion
As mentioned above, organic promotion is no longer going to be enough. SEOs have been used to the idea that link building / outreach is enough and while it is still effective, combining this organic promotion with paid promotion can give content the traction it needs to get in front of people. Our clients need multiple traffic sources in order to break through and this means a combination of organic, paid, social, email, direct and referral traffic.
5 - Our content needs to be beautiful
In order to break through the filters and stand out from everyone else, our content needs to be beautiful. I genuinely think that over the next few years, exceptional designers will be harder to come by than exceptional developers. One final statistic to back this up courtesy again of Mary Meeker:
Notice how the biggest turnaround in usage is from Instagram and Snapchat? I think that is because the next generation of internet users are used to the idea that content is visual in nature and naturally prefer to consume content like this.
6 - Learn what makes a good idea
Not just what makes a good content idea. We need to go deeper and learn what makes a good idea and this can improve our creativity. I'd highly recommend these three books on the topic:
I'd also recommend taking a look at this slidedeck from Mark Johnstone:
7 - Design and communicate mobile first
