|
A client of mine often expressed skepticism when others seemed to think that clickthroughs were sacred. And he was talking about all clickthroughs, whether those of people clicking through to his online magazine, or those responding to advertiser’s ads.
He’d argue that raw clicks might be unqualified people -- the top of the funnel, unfiltered visits that might not really be valuable visitors. He feared getting more traffic of that sort might be like paying
for a good meal and getting “empty calories.”
I pointed out that if we’re going to be skeptical of “top of funnel” unfiltered-quality response, we could take it all the way and ask a bunch more empty-calorie questions about whether clicks lead to: reading,
subscribing, re-visiting, word-of-mouth, clicking on your advertiser’s banners, reading advertiser site pages, responding to advertiser, buying from advertiser, repeat buying from advertiser, and word-of-mouth about
advertiser.
Yet clickthrough rates are what we can easily and quickly measure.
And if we want to get a fast learning curve going, we know it’s sometimes not feasible or wise to delay all interpretation until the whole funnel is analyzed and under control.
I’ve spent a lot of my
career on one side of this issue -- getting clients to continuously develop and measure more of the whole system, to maximize the end result.
Nevertheless, I intend to focus here primarily on the top
of the funnel, and to justify your taking incremental clicks more seriously (under certain conditions).
By implication, this will suggest why advertisers also take seriously those metrics that are just below
the top of the funnel -- things such as pageviews.
My client’s concern was based on the prevailing notion that techniques that produce more clicks will necessarily produce less valuable, less qualified ones. I want to challenge that assumption.
Notice that the very opposite could be true. Improved grabbers, treatment, placement, etc. -- the things we use to multiply click rate -- could actually yield even better qualifieds.
In fact, in all my
discussions with clients, and all my workshop sessions for their content and design teams, I’ve never uttered a single word advocating trading quality of response for quantity of response.
Instead, I’ve taught techniques for tuning in to the target audience’s deepest needs -- showing how that is the key to multiplying response while maintaining or increasing quality.
For example, consider the technique of writing great grabbers. A great grabber is not based on blind curiosity that would attract any human just based on provocative words and
pictures. Instead, such grabbers list and dramatize very specifically the value offered. In so doing such grabbers pull qualifieds like one pole of a magnet, and usually push away unqualifieds like the other pole!
Here’s another example. Consider an on-line content site such as NetscapeWorld, written for Web professionals. Obviously, the more specialized a technical magazine -- the more matched to the magazine’s
editorial will be most of the marketers who choose to advertise in it.
Then add the magazine’s commitment to using techniques and editorial content that pulls quality visitors, and what do their advertisers get? They get the assurance that clicks from that niche content site are likely to have far fewer empty calories than clicks from elsewhere.
Although I’ve never argued that content sites and their advertisers should compromise quality of average clicks to get volume, in what follows I’ll show that under certain conditions you might, indeed, want
to increase the absolute volume of qualified visitors -- a lot -- and in so doing allow a quality vs. volume tradeoff.
I hesitate to use “tradeoff” because as some of the scenarios below show, the “tradeoff”
can be entirely unfavorable, or entirely favorable -- not just gaining one thing by giving up another.
If the numbers below put you off, just read the narrative.
It will still make sense. But I hope you take the time to bear down on the numbers below, because the whole point is in how the numbers can change the game -- forming a tradeoff that qualitative, metaphorical thinking just doesn’t contemplate.
A few definitions will help make the following scenarios clear:
Let T1 = Total results (e.g. clicks) under old approach
q1 = Fraction of clicks qualified under old approach
Q1 = Total qualifieds (T1 x q1) under old approach
Let T2 = Total results (e.g. clicks) under new approach
q2 = Fraction of clicks qualified under new approach
Q2 = Total qualifieds (T2 x q2) under new approach
Now, let’s look at a some of the most interesting scenarios.
All but the last scenario have the “old approach” numbers the same --
as a common reference for the “new approach” that’s different in each scenario.
Empty Calories Scenario
First, here’s my interpretation of my client’s “empty calories” scenario.
T1 q1 Q1 M T2 q2 Q2 Q2/Q1
100 .8 80 2 200 .4 80 1.00
The numbers depict
this:
Under an old approach, of each 100 clicks, .8 are qualified, yielding 80 qualifieds. Suppose some new approach multiplies clicks by 2 times, but in such a way that the clickers now only contain .4 qualified. The new approach’s total qualifieds are thus 80 -- the same as under the old approach -- so Q2/Q1 = 1.0.
There’s no gain in qualifieds despite the doubling of clicks. The gain in clicks sends a false signal to advertisers, and incurs some costs (although quite low, given internet economics) for processing
the wasted page views of the unqualifieds.
I don’t want to get off my main argument, but put a pin in the following thoughts about word-of-mouth effects, and repeat exposure effects:
Even the above “no-gain” scenario, depending on the approach, might have a delayed gain if the unqualifieds are word-of-mouth transmitters to qualifieds, who later become more likely to click because they “heard about it.” Similarly, the unqualifieds might be worth something because they might later become qualified as their needs and other circumstances change. Such effects can be measured, but it’s hard.
Ignoring delayed gains, this “empty calorie” scenario seems pretty bad. But before I go to the paradoxical good scenarios, let’s notice that there are far worse scenarios than “empty.”
Poison Calories Scenario
T1 q1 Q1 M T2 q2 Q2 Q2/Q1
100 .8 80 2 200 .3 60 .75
This goes beyond empty calories.
Here the new approach still has a big 2x multiplier of clicks, but does so by sending a message that turns off the qualifieds (or exposes an audience with too few qualifieds), making the new fraction qualified even worse, and resulting in a loss in net qualifieds. This is a harmful “tradeoff.”
Poison Screening Scenario
T1 q1 Q1 M T2 q2 Q2 Q2/Q1
100 .8 80 .7 70 .9 63 .79
Above, in the name of improving the old
.8 qualified, an approach is tried that indeed improves to a new level of .9 qualified.
But in refining the targeting, the grabbers, etc. so they only grab those who are qualified, the multiplier goes down from an amplifier of 2.0 to an attenuator of .7. The result is to lose net qualifieds, bringing Q2/Q1 down to .79
Here you can deliver a higher percent qualifieds, but at the expense of how many net, absolute qualifieds, you deliver. This strikes me as equally poisonous.
It seems worth some effort to convince advertisers to not be impressed with such “screening,” and instead to notice the sacrifice of what really counts -- having Q2/Q1 > 1.0.
On the Internet, selling via
ecommerce, we can pretty much ignore the cost savings here from not having to process page views from unqualifieds. In traditional marketing, inquiry-handling labor, phone, and postage would be important, and
the whole argument about this scenario would change.
Having taken my client’s “empty” qualms into the “poison” realm, now comes the fun -- showing the paradoxical scenarios where multiplier approaches need
not be “empty” at all, if we look at net-qualifieds.
Nectar Multiplier Scenario
T1 q1 Q1 M T2 q2 Q2 Q2/Q1
100 .8 80 2.0 200 .9 180 2.25
This is the scenario I always strive for. Here, the new approach multiplies clicks, and does so in a way to improve the fraction qualified, in the above scenario moving up from .8 to .9. The net gain yields a Q2/Q1 of 2.25 fold!
I discussed how this is possible at the beginning of this article. For example, great grabbers not only multiply response, they’re often even better at attracting qualifieds while screening out unqualifieds.
I cannot say too strongly: The ratio q2/q1 can be greater than 1.0 -- even when one tremendously multiplies response. It depends on approach. Most designers of ads, and media planners, don’t know how to make everything better at once, which leads to the false assumption that there’s some inherent unfavorable tradeoff -- that q2/q1 must be less than 1.0.
Also, having a big Q2/Q1 like 2.0 doesn’t mean that we can’t achieve similar impressive gains as we go into additional stages of analysis, strategy, and creative. We can move through many cycles of
multiplying qualified results, as we move to Q3/Q2, Q4/Q3, etc.
Unlike cost-reduction programs which have a logical lower limit of zero cost, response-improvement programs have no upper limit -- except
market saturation. And even market saturation can often be repealed by promoting new offerings.
Another way saturation is avoided is the natural consequence of accumulating users and moving into the center
of the adoption curve, where persuasion economics become more favorable. At that point, you can convince previous unqualifieds to become interested and active. They’ve become qualifieds.
Pure Multiplier Scenario
A variant on the prior scenario is where the fraction qualified stays the same, making the results multiple the same as
the Q2/Q1 multiple.
T1 q1 Q1 M T2 q2 Q2 Q2/Q1
100 .8 80 2.0 200 .8 160 2.0
Tradeoff Multiplier Scenario
Now let’s look at a more paradoxical situation, the scenario whose great value isn’t contemplated by the usual
metaphorical thinking:
T1 q1 Q1 M T2 q2 Q2 Q2/Q1
100 .8 80 2.0 200 .7 140 1.75
Here the 2.0 multiplier is using an approach that does cause a
fractional q2/q1, but the “harm done” is less than the “good done” by the high 2.0 multiplier. The net gain, expressed in Q2/Q1, is an excellent 1.75.
How might this tradeoff happen? It might be
by adding some blind curiosity to the explicit learning offer, or by broadening the grabber’s interest, or by using less focused advertising media.
Given the unique economics of the Internet, allowing us to
virtually ignore pageview costs, this attenuation in q2/q1 should not concern the advertiser, as they should focus on Q2/Q1 instead.
The Internet offers another advantage in this situation over conventional
media, which is that the unqualifieds are invisible to the qualifieds -- so you don’t have to worry that the qualifieds will think, “if it’s for them, it can’t be for me” -- as might happen at a face-to-face group
event.
Sure, in this scenario there are some “empty calories” (carried in the wake of the big gain in net-qualifieds) but they will self-disqualify, and don’t matter. As discussed earlier, they may even
give word-of-mouth inputs to qualifieds, ideally saying “it’s too advanced for me, but looked like your kind of thing.” Or, the unqualifieds might later become qualified themselves, making exposing them not as empty
as it seemed.
But even if the unqualifieds are totally waste -- and assuming you don’t offend them -- encouraging them to click shouldn’t have any significant costs or risks. Certainly not enough
to forego, as in this scenario, 1.75x more qualifieds.
Low Initial Qualifieds Scenario
T1 q1 Q1 M T2 q2 Q2 Q2/Q1
100 .6 60 2.0 200 .5 100 1.67
None of the reasoning so far depends on the actual level of
q1. It only depends on the ratio q2/q1, and the value of M.
It can be shown that Q2/Q1 = M (q2/q1).
Simply put, whenever you can amplify total response more than you attenuate fraction
qualifieds in the process, you have amplified net qualifieds.
So in the above scenario, with a far lower q1 than past scenarios, we again have a very favorable tradeoff-multiplier scenario.
The limit
on this thinking is likely to be the point when an advertiser not only incurs the tiny cost of pageviews, but begins to process enough e-mail and phone inquiries to make the fraction unqualifieds be a real expense
burden.
But I must say, if these highly favorable tradeoffs can be had with negligible pageview costs, one can solve the inquiry expense problem better ways.
For example, the design of the advertiser’s site receiving the clicks can be improved to do a better job of signaling who should respond and why -- thus suppressing unwanted e-mail and phone inquiries, and their inquiry-handling labor.
Getting Off the Metaphors
So, “Empty calories” is a straw man.
As I’ve shown in the above scenarios, even big multipliers need not attenuate quality (and might even increase it), and it takes a big attenuation in average quality to lose the gains in net-qualifieds resulting from a big response multiplier.
I think you lose such insights and fall into false conclusions if you think about these issues in terms of metaphors. And such metaphors abound...
One friend likes to talk of the “wasted
lives from sluicing for gold where there isn’t much, where you ought to try and find a rich vein and use a pick and shovel.”
He also talks of “the folly of giving away ice cream at a trade show booth,
yielding thousands of irrelevant visitors and ‘prospect’ names.”
All those metaphors may be good persuasion if dealing with someone else’s, or one’s own, temptation to gather response for the sake of
response, while ignoring quality and effort. But metaphors can block one’s mind from thinking about how to create everything-improves type multipliers, or more-net-qualifieds tradeoff multipliers.
The Empty-Calories Paradoxes
As the scenarios above have shown, we have the following paradoxes, that become true if you have
good enough Persuasion Design underlying your multipliers, and the Internet’s low-cost of processing visitors:
- There’s something far worse than empty calories: Poison Calories Scenario.
- Screening and niche approaches can increase the fraction of qualifieds, but attenuate net qualifieds: Poison Screening Scenario
- Incremental clicks can be of higher, not lower quality: Nectar Multiplier Scenario
- You can accept a lower percentage of unqualifieds, and yet gain net qualifieds: Tradeoff Multiplier Scenario
- Many cycles of improvement may be possible with net gain in qualifieds each cycle.
- It would be a negative gratuitous assumption to assume that multipliers reduce net qualifieds.
Therefore, increased clicks can be a useful metric after all, correlating highly with increased net-qualifieds, and giving a valuable immediate reading. Likewise for other early measures in the funnel, like pageviews.
I hope you find this useful as you consider techniques for increasing clicks into your site, or within your site -- or as you plan ways to increase clickthroughs from your ads. Email or phone me if you have comments or questions about any aspect.
|
