The six immutable laws of digital marketing
(Note: The six graphics that accompany this article have been placed on a separate web page to facilitate easier reading. Click here to open the graphics in a separate window.)
Although the idea of immutable laws of marketing is not new and is explored in detail by Al Ries and Jack Trout in their book ‘The 22 Immutable Laws of Marketing’, a quick search on Google reveals little in the way of the possible existence of immutable laws in digital marketing.
Given that digital marketing is so accountable and generates substantial amounts of data, there would seem to be a particularly good opportunity to use basic statistical and mathematical techniques to develop models that help to explain the variations digital marketers see in their observed events and data.
Perhaps it is these models that we could consider to be the laws, rather like other notable laws such as Moore’s Law, which deals with the density of semiconductor circuits.
A law that is less familiar to most, but potentially more relevant to digital marketing, is Zipf’s Law. Derived from linguistics, Zipf’s law has relevance for search engine campaigns and online shopping web sites. This is discussed more fully in an interesting article by Dave Chaffey.
The essence of the application of the law in search engine marketing is that customers start their search broad and then get increasingly specific as they narrow their search. A good example is a web searching or online shopping session that starts with ‘gifts’, then selects ‘birthday gifts’, and finally ‘birthday gifts men’.
So there are at least one or maybe two laws that have relevance to digital marketers — but that just doesn’t seem enough. I propose that there are six other possible laws of digital marketing. These I have characterised as:
- The Slow Build.
- The Spike.
- The Ceiling.
- The Exponential Growth.
- The Gestation Period.
- The Straight Line.
I have included in the following explanations visual representations of statistical analyses of real data. Where relevant a mathematical model, including formula and R-squared value, is shown to indicate how well each model explains the variations in the observed data.
When we think of Christmas shopping, we think of the Christmas shopping rush. Online, however, Figure 1 shows the trend in click rates on shopping-related search terms in the days before Christmas.
What is evident is what I call ‘The Slow Build Law of Digital Marketing’ — versus the mad rush! Sure, there is a bit of swing because of normal intra-week variations, but the law is explained well by the mathematical model.
The slow build effect is also seen in other seasonal trends such as the school holidays. As the recent July school holidays progressed, there was a slow build of searches for ‘what’s on melbourne’, presumably from parents with school-aged children.
The positive thing about the law of slow builds is that marketers have time to take advantage of a trend, which typically lasts for a period of weeks, unlike the spike, which as we will see is almost over before it has begun.
The spike is the most transient of digital marketing effects and lasts for no more than a few hours. During this time a sudden rush of visitors descend on a web site, rather like opening the doors at the Myer stocktake sale.
Although there is a positive side in achieving a large increase in visitors to a web site, a rush can cause a headache for IT departments and server load capacities as well as for customers themselves.
Many of us have experienced the ‘server unavailable’ message when tickets go on sale after special promotions are released through advertising in main media.
The spike shown in Figure 2 is the website response to the promotion of a competition during a segment on Channel 7’s Sunrise program.
This particular spike generated a more than 800% increase in average visitations at its peak. However, visitation dropped off as quickly as it grew and the residual level of visitation returned to almost normal levels.
Spikes can be expensive to generate as they require an investment in main media vehicles, typically TV or radio, although sometimes the same effect can be seen as a result of some heavy internet media advertising or a good piece of viral marketing.
‘The Ceiling Law of Digital Marketing’ states that if you email your database you will never get 100% of customers opening their emails — even months or years later. In this example (see Figure 3), a client purchased a sponsorship of an email newsletter on the Fairfax Digital network. The curve represents the ‘opening’ of the emails in the days after the email was broadcast.
With around 50% opening their emails within a day of broadcast, digital marketers have delivered their message to the bulk of the target audience within a couple of days.
With the very long tail, you should also work on the assumption that the ads may still be unread and in a customer’s inbox in 365 days’ time, thus the promotion may be well and truly past its use-by date — if it gets opened at all.
Figure 3 shows the cumulative proportion of emails opened by subscribers each day after the email was broadcast. Interestingly, we would see a similar phenomenon (commonly called an audience reach curve) in mainstream media, particularly where audience reach and ratings points are concerned.
In principle, it is relatively efficient to achieve 50% to 70% audience reach quickly, but practically impossible and prohibitively expensive to achieve 100%.
Law 4: The Exponential Growth
(see Figure 4, separate graphics page)
You might intuitively conclude that the growth of the number of broadband connections in Australia might resemble Moore’s Law, driven by ever-faster computer chip speeds.
Upon closer examination of the number of broadband connections reported in the ACCC’s Snapshot of Broadband Deployment, the ‘Exponential Growth Law of Digital Marketing’ offers a better explanation, as shown in Figure 4.
With the model explaining 98% of variation, the only other time I have seen a curve like this was in high school biology while studying the rate at which bacteria grew!
The exponential growth law is really about market diffusion, of enabling internet services that enable digital consumers to consume and digital marketers to advertise. As we have heard many times, advertising dollars follow eyeballs or, in this case, connections.
‘The Gestation Period Law of Digital Marketing’ is simply the time (measured in days) between the first web site visit by a prospective customer and that customer making an online purchase.
Based on an analysis of first visit and purchase intervals for customers purchasing a consumer telecommunications service online, up to 63% of customers who purchased did so within the first 24 hours of visiting the web site. The remaining 40% purchased the product sometimes one plus days later, as shown in Figure 5.
Therefore, even though you may stop your digital marketing efforts at a certain point in time, the momentum will continue to deliver sales for some time beyond! I guess the trick is not to stop marketing.
It is somewhat fitting that ‘The Straight Line Law of Digitial Marketing’ is the simplest yet most profitable. This law is a characteristic of the online medium owned by the largest media company on the planet — Google.
Although the Google bidding formula takes into account both click rate and bid price in calculating a ranking, my own analysis of real campaign data is that over the long term bid price appears as the key driver of listing position.
This occurs because the long-term average click rate is almost a constant somewhere between 2% and 4%, depending on your product/service (even taking into account seasonal variations such as I discussed earlier).
Based on real data for ‘broadband’-related search terms, paying 10 cents per click will get you somewhere between positions 4-5, while paying 35 times more (i.e. $3.50) will get you into number 1 spot.
What needs to be weighed up at this point is whether the rank 1 spot will return the highest level of sales. Based on research from ATLASDMT, it appears that the lower the rank (i.e. higher the position), the better the sales conversion rate.
The other interesting point about this law is that the left-to-right slope of the line will get less steep if advertisers increase their bid price to maintain control of the number 1 ranking. In other words, this law also forecasts increased per-click revenue for Google!
Conclusion
This is not a formal lesson in statistics or mathematics, as I am neither a teacher nor an expert in statistics. What is interesting from my perspective is that with the help of some simple statistical models, digital marketers can characterise digital marketing activities using readily available data.
Hopefully, in the process, you can gain a better understanding of some common digital marketing response curves and naturally occurring market dynamics that will hold true time and time again.
The six immutable laws may be of assistance in anticipating the sort of visitation, exposure and purchase trends you wish to generate with your digital marketing and media activities.
So to the creators of the Big Ad, www.bigad.com.au, which of the six laws of digital marketing did you anticipate? Was it the spike or, as it should be more appropriately named, the Burp?
|
