Fundamental to creating any digital marketing strategy is a firm understanding of what it is that channels do and who they can reach.
This sounds straightforward, yet I regularly see companies that are reliant on channels that represent a race to the bottom (in terms of price/margin) or who are paying more than once to convert a customer. This is often a sign that they are mainly targeting low-value customers and/or even their own existing customers.
I’ll discuss the issue in broad terms for the purposes of this article. Of course, there are exceptions to almost every argument but that shouldn’t diminish from the overall point.
Before optimising any advertising, the strategy must make sense. For it to make sense, channel capability must be understood. What’s channel capability, you ask?
Customers can be broken down into three types:
a) Existing or lapsed customers – They have bought, are considering or are going to buy your product or service.
b) Competitor’s customers - They’ve bought or are going to buy your competitor’s product or service.
c) Potential new customers – They have not yet bought from either you or your competitor, they haven’t heard of you. They now have or will have a need for a service/product like yours and may or may not know it.
The most valuable customers tend to be the ones which are the newest to the market. If you can reach them at this stage, they’ll pay more on average and be more loyal.
Competitor’s customers can be enticed away but you’ll need to do something special. Rare companies have a brand, product or service that will sell itself, but the reality is that for most, they’re accepting a lower margin to get the customer. This can work if you can keep them long enough but typically, you’re attracting the sort of customer who’ll continue to look around for something better – and by that, I mean cheaper.
Your existing customers are a goldmine and you need to focus the right channels on extracting that gold. Your existing customers do not need additional advertising and neither do people who have already decided to buy from you. In the main, brand PPC and re-targeting are exercises in reducing margin and ROI. Perhaps not always, but if you think they are or might help, make sure you plan thorough testing to prove it.
Non-brand search (SEO & PPC)
Non-brand search is where the world airs its dirty laundry. Used effectively, it’s a database of customer intentions and both SEO and to a lesser (but still huge) extent PPC allows you to tap into it. You can figure out what potential customers might be searching for and get your content seen in those areas. What might a new customer who has never bought before search for? Write content that answers that need and you’ve got a chance of making that first connection with someone who might become a customer.
PPC allows you to pay to be visible anywhere that SEO does. It also has an advantage over other forms of advertising that search engines disguise it so up to 60% of people don’t know they are clicking an advert. If the brand is happy with this sleight of hand, there’s a lot of opportunities there. Typically, brands will focus their spend on big money keywords which mainly target competitors’ or their own customers. Experimenting with lower CPC terms that bolster visibility at the top-of-funnel can provide better lifetime value returns.
Content marketing is such a wide and varied channel, it’s hard to give any sort of overview. It is fair to say that when done well, it is possible to get fantastic and useful content in front of potential new customers who have never heard of you before. The key is to measure performance and this is almost never simple last click conversion.
Display is either completely ineffective or (at best) a reminder to people who were already planning to become a customer. Often, your ads are not seen or didn’t even happen at all. If you’re lucky, they get seen but ignored.
As a generalisation, display is favoured by large, old school media buying agencies. The language and function are more similar to print, TV and so on. There’s almost as little transparency attached. Programmatic advertising was considered as a great hope for display but ad fraud, a complex and opaque ecosystem and inability to show positive results have all but killed it for serious marketers. The exception is re-targeting which is the ‘reminder’ I mentioned before; showing your ad to people who are already aware of your brand, including everyone who has already decided to purchase. Identifying what effect this has is not straightforward and needs AB testing.
Years ago, affiliates were explained to me and I thought it sounded like a great marketing opportunity: you pay websites a commission for any conversion they drive. This incentivises them to promote your service or products.
In theory, this sounds great, you’re almost guaranteed to make money, you don’t need to lift a finger and you only pay when you make sales. Too good to be true? Yes. In reality, most affiliate sales are comparison sites like Moneysupermarket, Skyscanner etc. or voucher sites like Quidco etc. What these do (at best) is give you the edge over a competitor, usually by cutting your margin in one way or another - often you’re also paying fees and cutting prices on sales you’d have received anyway. In short, affiliates typically get you low-value sales and reduce your margin in other channels too. In many verticals they are dominant and it can be hard to do without them.
Social media advertising
Social media advertising has claimed similar legitimacy to paid search in recent years but the reality is that normally it’s not an efficient way of reaching new customers. It is much better at reaching the sort of people who have already decided to purchase from you or your competitors. Ten times better, to be precise, according to experiments carried out by Facebook. The ratio of advertising effect against selection effect was found to be 1:10 meaning for every ten people who purchase after clicking your ad, nine would have purchased anyway. This makes a huge difference when calculating ROI. If I spend £10 on ads and sell 10 x £10 widgets, my ROI is 1:10. If nine of those sales would have happened anyway, I’ve spent £10 to make £10 revenue.
This figure varies significantly in Facebooks experiments which confirms how well you need to understand what’s really going on in order to work out what return you’re getting.
Social media is an interesting channel in that different people often mean completely different things using the term. In some ways it can be an exciting way to get eyeballs on your brand and doing so could bring you new customers or increase sales. Paddy Power and Burger King are good examples of creative and brave brands that are able to use it well. However, 99.9% of the time, Social media for brands means stale corporate social media accounts with nothing to say. The accounts are targets for disgruntled customers and usually ineffective at handling them. The bottom line is that for most companies, social media accounts are followed only by existing customers, randoms, and bots. They are not able to reach and convert new customers.
Brand paid search
People who search on your brand already know who you are. It is a navigational search – they simply want to get to your site. There are legitimate uses of brand PPC but leaving it on all the time is not one.
What might appear to be criticism of the channels above is not. The intention is to highlight where, through assumption or being misled, marketers think that certain channels can reach types of people that they typically don’t. This leads to bad strategy and wasted budget.
I’m not suggesting you don’t use any of these channels but to use them effectively, you’ve got to understand what they do, who they can reach, how to test them and how to measure success.
It might be tempting to look at your report which says affiliates cost x and brought in 20x and leave it at that but you’d be ignoring both the loss of margin in other areas and the wasted affiliate fees. You could spend a ton of money on a great social media strategy and be happy that it looks nice and brings in some sales, but you might be ignoring that the sales were from existing customers who were most likely to renew anyway.
I’m aware that the ‘how’ of measurement is not answered in this article. The short answer is it involves AB testing, measuring lifetime value, and effective channel attribution. I’ll leave the detail for next time.
Richard Falconer is managing director at Yard.