If you don’t use Google Insights much, I’d seriously recommend adding it to your day-to-day strategy toolkit – regardless of whether you are a digital specialist or not.
In most categories, whatever you communicate and wherever you do it, at some point, your potential customer is probably visiting Google. You can use Google’s other free tool Adwords to work out how they are researching you and your competitors.
For instance, Google tells us that in September 2009, Australia searched for:
- ‘tampon’ 22,000 times
- ‘bunk beds’ 135,000 times
- ‘camera’ 2.74M times
- ‘new car’ 1M times
- ‘pencil’ 165,000 times
- ‘cereal’ 90,500 times
There are few perfect data sets around so I use the examples above purely to make a point I frequently make to marketers who typically have 2 big sticking points when it comes to digital:
1. My target audience isn’t online
2. They’re not interested in what we’re about online
To address the first issue – I do 2 things. First, I ask the marketers about their own digital activity, whether their IT department allows them to access YouTube, Facebook… and whether this may impact their perspective on ‘everyone else’. Secondly, I’ll dig up some Roy Morgan stats, some Nielsen stuff. There are plenty of stats online, on newspaper websites. I have some in my social media presentations on Slideshare.
One thing I try not to do is blindly evangelise digital. ‘Show don’t tell’. Show other people’s credible statistics to make your point. Also, it’s not just about digital. Doing stuff offline is extremely useful in driving search behaviour… online. There is plenty of research that tries to identify people’s behaviour after they see/hear communication offline – do they remember or use a URL, do they Google? Typically, Google is first port of call… even if your URL is written big.
To address the second issue – I make 2 points. First, I’ll show search stats like the ones above. Numbers are very powerful when you’re talking to numbers people. I’ll also organise the keywords into groups according to a path to purchase (from broad, awareness-type words – like ‘cameras’, to tighter phrases like ’sony digital cameras’ through to ‘best price [model x]…’).
I’ll then make the point that part of what a company needs to do online should address the path to purchase stuff – help people make informed decisions about buying… BUT if your brand has a broader purpose – such as helping people maintain a healthy weight – I’d then do some more keyword research and find words like ‘fitness’ and ‘weight’ (both with 1.83M searches each in September), ‘loss weight’ (823,000 searches) and ‘weight management’ (90,500 searches) – and do some quick maths and see that between what your product is about, and the bigger topics you can help people with that the numbers can quickly look… well, large.
A quick look at the Google impact of 7 non-traditional Australian campaigns
1. Vegemite invites Australia to name its new product – 2009
Vegemite is a quintessential Australian food (don’t get snarky!). Recently, they invited Australia to name a new snack-styled variant (read commentary at Mumbrella).
The blue line above shows searches for Vegemite in Australia over the past 4 years. There was a strong peak in 2006. And it’s going crazy now off the back of the new campaign and all the media commentary. How long this peak stays high, whether it converts into sales… we’ll soon see.
Interestingly enough, iSnack is also getting a lot of search activity.
Since Google Insights only shows a comparative index of search volume, if you have time, you could jump into Adwords to find rough numbers to get more academic with it.
Thoughts: Using Google data for Vegemite in this situation will only tell part of the story. They obviously have Australia’s attention; some of the attention would be converting into sales but it’s really now up to them to turn the attention into something more tangible over the long term or they’ll risk being a ‘one hit wonder’ in the digital age.
2. VB beer, Warnie and Boony
‘VB’ has a tonne of searches but they’re for a bunch of different stuff (eg ‘VB script’) – not just beer. So I decided to look at ‘VB beer’, ‘VB warnie‘ (Shane Warne campaign) and ‘VB Boony‘ (Boon campaign).
What’s interesting above is that ‘VB beer‘ has really declined from 2006. Perhaps more competitors entered the market? Perhaps search behaviour changed because relevant results weren’t being displayed? Yes, I’m sure many would say that the VB audience isn’t exactly Googling the brand, but I’d suggest this is too simplistic a dismissal. All I’d do is explore why this may have happened. Search is often a leading indicator of interest in a brand.
‘VB warnie’ really captured our attention in 2007. But search for ‘VB beer’ declined through 2008. Does this match brand tracking results? Perhaps the Boony and Warnie campaigns worked from a sales point of view but proved to be gimmicks that did not change long-term perceptions of the brand? I don’t know – it needs exploration and access to a lot of other data sets – but you can definitely see a spike in searches for ‘VB beer’ after the new Droga5 campaign from mid-2009. Does this spike match movement from brand tracking?
Thoughts: I’m sure it’s been done, but the first thing I’d be looking into based on the above is the correlation between promotional activity, sales, Google activity and brand tracking. Perhaps the sales promotions that became so famous didn’t actually make VB more famous, more liked, more consumed?
3. Sunsilk National Hair Survey- 2005
I spent a few weekends well into the night working on the technical implementation of this campaign at MassMedia (read case study). I remember it well! Leo Burnett/arc (creative) and Universal McCann (media) put the strategy together.
The website featured a 20-minute survey about hair. I admit that I didn’t think many people would fill it in… despite the cash incentive ($25k). Over 250,000 people did.
If you look at the graph above, SunSilk was declining pre-2005, had a massive spike at the time of the campaign, kept fighting but has declined quickly since.
Thoughts: My gut feeling on this is that sales promotions – as much as they get you in with Coles and Woolworths – cannot be your only solution. If you have a big brand problem – a sales promotion is a great way to get attention but your real opportunity is to actually change with the market. SunSilk tried to do this by making their products more ‘easily navigable’ (‘if you have hair type A, use product X’) but this felt like a shallow change… this is just a hunch. Without sales and brand tracking data, it’s hard to make a conclusive argument.
4. McDonald’s and Name It Burger
To give you a bit of context, the word ‘mcdonalds’ is searched about 550K times per month in Australia. If I was McDonald’s CEO, I’d like the graph above. The brand seems healthy. People are interested and want to find out more.
You can see a little spike around the time of Name It Burger (read case study) for both the product and the brand.
Thoughts: The Name It Burger promotion played a small part in McDonald’s bigger picture action – revised menus, revitalised restaurants, more premium advertising. Their strategy has been holistic and it looks like people are liking it. Again, match this data to brand tracking and sales data for truer picture.
5. WWF’s Earth Hour
The red line above clearly shows that WWF is declining. Yes, there are various organisations with the same acronym but a quick look at Adwords didn’t help much in pulling apart the World Wildlife Fund search behaviour.
The blue spikes show that 2008 was Earth Hour’s most impactful and that the spike in searches was quite compact. The highest web activity – from memory – was on the day of Earth Hour and within 2-3 days of either side. Search activity coincided with this.
Thoughts: Earth Hour wasn’t about WWF getting better known. It was more about getting people to act, to put politicians on notice – so I wouldn’t be as concerned about the failure of Earth Hour to also push WWF searches up. However, I would be worried about the general decline. There’s more competition in the not-for-profit and donation space than ever before. Perhaps it’s time for WWF to re-assert its brand.
6. Dove Real Beauty
Like the SunSilk National Hair Survey, I was involved in project managing the digital implementation of this in Australia.
Dove was a brand in decline when the Campaign for Real Beauty launched around the world. There’s a small bump in the graph above in 2006 – it coincides with a solid spike for searches for ‘Dove’. Since that time though, I can’t recall Dove doing much locally and the search traffic has declined accordingly.
Gut feeling: I see the above graph and see a brand with a very insightful strategy, trying to do and be something good. Deep down, I wonder whether the almost-duopolistic nature of the Australian supermarket industry distracted the local Dove team from continuing to build something even bigger and more amazing, resorting, instead, to discounts, decals, shelving at the end of aisles.
7. Tourism Queensland – Best Job in the World
The most alarming thing about the above graph is the decline in searches for ‘queensland’ since 2004. Now, this graph alone could actually be mis-leading. Perhaps people’s search behaviour changed. An example: one client I work with is mandated to use a global website with one of those landing pages that shuffles people through in a way that only makes sense to people at the company – not their customers. So a lot of search behaviour starts with their brand name – but people hit this bad user experience so they then refine the terms they use to find a better, more local website.
Anyway, look at the red line above for ‘best job in the world’, it definitely got some interest but hasn’t managed to change search behaviour for ‘queensland’.
Gut feeling: Perhaps, like the VB Boony and Warnie campaigns, was Best Job in the World a famous campaign that did not change behaviour and lead to more sales? Touch call. To really understand this, you’d need to look at category data (ie travel overall, travel to Australia overall) and a much more detailed search analysis.
Above is the search behaviour over the past 12 months. Not much else to point out really. ‘Queensland’ is still in decline.
What to take out of this
1. Search behaviour and metrics should be a key business metric – for CEOs and marketing directors
2. Search is a leading indicator in many categories about purchase intent
3. Use keywords to help refine your brand strategy – not just digital strategy
4. If you can mine search data, as well as brand, sales, survey, website data – well, you should be in a good position (but realise, that if you ONLY spend time in numbers you’ll miss the intuitive opportunities)
5. Sales promotions – like ‘viral campaigns’ – may get you quick wins, they may get your agency in the papers but are they actually changing behaviour? Don’t fall victim to short-term blinkers. Use these quick fixes AS you also unleash more holisitic changes to how you deal with customers, product innovation, and so on.
What do you think?
Photo courtesy dullhunk.
If you enjoyed the read, please leave a comment. Feel free to follow me on Twitter
