|
|
This edition of TMAI Premium is free for Standard Edition subscribers.
The last few weeks in Premiumlandia
have been super exciting. We obsessed about how to build storytelling skills, using
before and
after examples. The result was 18 high-impact rules.
The last two weeks, I've focused on a particularly difficult challenge: Analysts getting scapegoated (for bringing bad news). The newsletters shared:
A.
How to recognize scapegoating, and
B. 11 effective strategies to deal with it (my
contribution to your emotional well-being!).
You can upgrade to TMAI
Premium here. All Premium revenue is donated to charity.
|
|
|
TMAI #307: Got Low Incrementality? Six Causal Factors.
[
Web
Version ]
|
Attribution
is NOT incrementality.
Incrementality is the art and science of identifying
results that would not have occurred otherwise.
In the context of Marketing: What would we NOT get if we cease all Marketing?
TIP: Attribution is simply distribution of credit for incremental AND
non-incremental sales. When you want to trick your bosses about your impact on the business, use attributed revenue.
If you want to tell the truth, use incremental revenue.
[TMAI Premium members, see TMAI #233 and TMAI #260 for how to measure the true incrementality delivered by your marketing budgets. It is one of the, if not THE, hardest
thing to do in Marketing Analytics, but lord, it is so much fun!]
I’ve had the privilege of measuring incrementality across the companies I’ve worked, or consulted. I particularly love answering the incrementality question because I’m
answering it for the CFO of the company, and that delivers major influence/impact. (CMOs who want to know incrementality are rare, and to be cherished.)
When I present the results of measuring incrementality, there is always this crucial question:
Why is the incrementality of our marketing, so low?!!
Sometimes, the implication of that question is that the analytics are bad (they can be, but usually not).
Other times, this line of thinking is a prelude to an attack on the Analyst because the numbers make the leaders look bad (see TMAI #305, TMAI 306 for how to identify you are being scapegoated, and how to avoid being scapegoated!).
Occasionally, there’s genuine curiosity. This is the best scenario because the cultural posture is:
How can we do better?
This week, I want to outline six causal factors that help explain low incrementality from your marketing initiatives. My hope is this will help accelerate your incrementality to insights journey, and demonstrate the massively strategic
nature of the causes of poor incrementality. When addressed, you have an opportunity to deliver a transformative impact.
Ready?
Let’s go.
|
1. The Wrong Business Problem.
Let's say you want to increase Unaided Awareness of your product/service. It turns out that you already have 90% Unaided Awareness. That number indicates you are solving a problem not worth solving. Hence, when you spend money solving
that problem... There might be very little incrementality (or none at all!).
Or, you already have the #1 Organic search result for every brand term. Not only that, your SEO team is so good that you have each #1 result broken down into relevant sub-links, connecting keywords deep into the site, etc. Now, if your
Paid Search program is focused on only buying your top brand keyword queries by humans, it is possible that there is little incrementality identified when you look at your campaign results.
I pose the following two questions to ensure that Marketing is focused on initiatives that will drive incremental results:
A. Are you solving
a problem that's actually a problem?
B. Is that problem worth solving?
If you have poor incrementality from your marketing initiatives: There is a very high probability that you are not solving an actual problem (see examples above) OR you are solving a really small problem (small feature, small product,
few customers impacted by it, etc.).
Either way, rethink strategy.
If you want your marketing to deliver high incrementality, do work that's worth doing.
[Bonus lesson: From here on out, nearly all
why not more incrementality reasons will come down to A or B or A & B. When you mull deeply on the causal factors below, you'll see they have a strong relationship to
A or B or A & B.]
|
2. The Poor Success KPI.
I want to emphasize this is not an analytics problem, it is a strategy problem.
Let's say you have been a successful DTC eyeglasses company. Say, the wonderful Warby Parker. You've blitzed television with your ads, you've purchased all the ads on Instagram, and, best of all, you've built just the kind of quirky creative
that plants a memory hook.
Your sales have started to flatten out. So, the CMO says:
We need more people to know who we are! Let's spend the next $100 mil buying the rights to the NFL and combine that
with more TV commercials so that we can increase our Unaided Awareness. That will fix the sales problem because the more people who know us, the more chances they will buy from us. GO! GO!
Why will this have extremely poor incrementality?
Warby Parker’s Unaided Awareness currently stands at 69%! Clearly, their Marketing over the last decade did not suck!
There’s poor incrementality from this campaign because it is extremely expensive to move metrics that high up – the diminishing returns curve flattened a long time ago.
[Technically speaking: WP’s problem is that there is no
headroom. Obsess about identifying headroom for each KPI chosen.]
If the WP CMO dug a little deeper, she/he will discover that an opportunity for WP exists for Purchase Intent. The baseline for Purchase Intent is just 8%. Headroom!
If they proceed to spend $100 mil and measure it for Unaided Awareness, the campaign will show a failure.
Pick the correct success KPI. If the problem your marketing is solving is worth solving (#1 above), then pick the KPI that is clearly aligned with that problem
AND ensure there's headroom.
If you've maxed out the sales you will ever get through digital (no headroom), shift the KPI to solve for in-store visits (in the beautiful Warby Parker stores!). Your marketing campaign will suddenly show high incrementality.
[PS: To be super clear: All data about WP above is made up, to tell a story. I have no special access to their data. Love the brand.]
|
3. The Audience Misalignment.
Let’s say I work at Amazon, and I’ve just purchased Whole Foods (because why not) and Mr. Bezos has ordered me to take Whole Foods’ profit to the stratosphere.
I need to squeeze more juice out of the ethically sourced, sustainably farmed, gently massaged, beautifully groomed, buffed and shined to a high gloss, lovingly shelf-arranged
grapefruits.
My cheap and fast answer? We need more people of a particular ethnicity, with
fatter wallets, and higher education who feel responsible for the impact on the environment to shop at Whole Foods.
BOOM!
A $40 mil digital campaign hyper targeting the above psychographic in a 23-mile radius of every single Whole Foods in the United States. These campaigns have beautiful creative, delightful calls to action, and LOTS AND LOTS of
ethically sourced, sustainably farmed, gently massaged, beautifully groomed, buffed and shined to a high gloss, lovingly shelf-arranged grapefruits.
A month later, the Analyst comes back with: No brand lift in Purchase Intent. No statistically significant lift in same-store sales across the test/control cells set up for the campaign.
Great campaign. Well targeted. Fantfreakingtastic creative.
What do you mean failure?!!
Well. I picked the wrong audience.
Every person of a particular ethnicity with fatter wallets and higher education who feels responsible for the impact on the environment already shops
at Whole Foods – or has an excellent reason not to.
I should be solving for audiences who do not currently know of, or consider Whole Foods. Maybe it is those irrationally loyal to Trader Joe’s people. Maybe it is Asians who write successful newsletters on Analytics. Maybe it is people
who lack access to fresh produce in their neighborhoods. Maybe it is people who shop at Amazon but have never been to a Whole Foods. Maybe it is… Some. Other. Audience.
It is instinctive for Marketers who either, A. solve for people like themselves (the reason marketing in almost all large companies largely
fails in the real-world), or B. their existing audience.
Both glorious paths to failure. Sorry. I mean, both glorious paths to low incrementality.
Your TikTok campaigns targeting the youth are delivering diddly squat because only old people are on TikTok now. The kids are in Discord (good luck reaching them there) or soon enough (lord forgive me) the Metaverse!
Consider the Cost Per Incremental Sale for Oculus TV commercials targeting THE entire United States vs. campaigns for Oculus that could perhaps have their target audience evaluated with a bit more focus and care.
If you want incrementality from your Brand or Performance campaigns on Television, do you need anyone other than Sports Fans? Then, look elsewhere. 98% (yes, ninety-freaking-eight percent!) of the most viewed broadcast programs were sports.
That number is still a very high 72% across cable programs.
Think that, and then checkout the last Salesforce ad in heavy rotation on TV with Matthew McConaughey in a balloon doing lord knows what for 57 seconds and in the last three seconds he says save the earth and screen fades to 0.4 seconds
of the Salesforce logo. Who is that ad for? And, if it is for all of humanity, is that an audience Salesforce should really solve for? I mean,
really?
If you pick the wrong audience – too wide, too narrow, too new, too old, too young, too current, too disconnected, too…. — anticipate low incrementality from your marketing.
Don’t be like me and my execution of my Whole Foods campaign.
[PS: Reminder: I do not work at Whole Foods. I do love shopping at Whole Foods. I do not work at Salesforce, and have no access to their strategy/data.]
|
4. The Terrible Creative.
That Super Bowl Salesforce ad on YT: 1,195,390 Views | 3.6k Likes.
Not exactly a ringing endorsement of the creative.
In our earlier TMAIs, we’ve learned the extraordinary impact creative has on the ultimate results (TMAI Premium #224) and how systematically analytics can help identify how to build a creative that delivers an extraordinary impact (TMAI
#265, specifically for YouTube).
Approx. 60% plus of the ultimate impact of a campaign is determined by the creative.
That Progressive ad on TV. The text you see on Bing Paid results. That (annoying? :) display ad you see on Tampa Bay Times. The interruptive in the middle of the online video you are watching.
That drives a bigger impact (incremental results!), than your targeting, platform, day parting, machine-learning powered real-time optimization, audience selection, media planning, frequency shaping, and the thousand more things that you
are doing to deliver business impact.
[Confession: I’m embarrassed to admit that for the first decade of my career - I’d already published my first book! - I only focused on the 40%, and was not even aware
of the 60%. I’ve worked very hard to make up for this mistake, but I continue to be embarrassed. Please don’t be like me.]
So, how good is your creative?
Oh, and how do you know that it is good and will deliver impact? If you know, you can fix or traffic more.
Is it built only for your company executives, or it is built for the actual audience the media is being delivered to? One of these two is the right choice.
Is the creative solving the problem that is being measured by your success KPI? (Such a massive issue, a place where good creative goes to die because you are
judging a fish by its ability to climb trees.)
How quickly can your current analytics implementation accumulate data on creative performance after your campaign is live – to learn if you suck, suck less, or rock? (See TMAI #273, the
win while you spend strategies.) #inflightoptimizationrocks
How consistent is your creative strategy? Changes concept/brand cues/values/talent/creative execution/formats every quarter? Every other quarter? Every year? Consistency delivers higher incrementality.
Is one campaign’s creative harmonious across channels, or do you have a different creative strategy on every channel? A silo strategy is less effective than a portfolio strategy.
I have a ton more to say on this topic. If I started my career again, I would grow up learning and getting better at
creative quant and qual analytics, than ERP and Digital analytics (the path I actually took).
Effective creative equals high incrementality. Period.
[Bonus reading: TMAI Premium #294: Why does so much advertising fail?]
|
I will cover the next two in detail in future Premium editions, Google Docs is warning me this newsletter is getting too long. :)
5. The Poor Media Strategy.
We started with one example in #1 of a poor media strategy. All you do is buy five brand keywords on Bing/Google/Seznam.
There are a hundred more ways to get your media strategy to be wrong.
Effective frequency on Facebook is 3. You buy 1 or 18. Neither will show incrementality.
You need 150 GRPs a week for 6 weeks to have a MINIMUM chance of success. You buy 148 GRPs for 5 weeks and six days - No incrementality for YOU!
Etc. Etc. Etc. Yada. Yada. Yada.
A hundred more ways for Media to deliver low incrementality. Watch for them. Hate them.
|
6. The Sucky Measurement.
It is possible you are showing your marketing delivers low incrementality, because you don’t know how to measure it correctly.
Unlikely, but it is possible.
Maybe you are one of those that is parading around with your attribution analytics reports built off a custom attribution model with partial media in a clean room. It shows low incrementality, and it is the measurement’s fault – wrong
tool, wrong answer.
Perhaps you don’t have the one advanced skills statistician with deep marketing experience, and you don’t know how to build models to assess incrementality.
Maybe you have one small incrementality initiative, but extrapolate it to global budgets/executions. Poor measurement.
Etc. Etc. Etc. Yada. Yada Yada.
Twenty-eight more ways to deliver wrong answers. Watch for them. Fix them.
|
Bottom line.
This is one thing I really, really, want you to get: When you report low incrementality, the first thing that is going to happen is you are going to get shot – as messengers often do.
[Did I already recommend you read my absolutely essential-reading-for-analysts newsletters on scapegoating, TMAI #305 and TMAI #306 and that they will change your life?]
What you’ll discover is that low incrementality is a result of everything except (usually) measurement.
It will take a lot of courage and influence to shift the focus to root causes (#1 through #5 above). Open-hearted cultures, where blaming is a tertiary concern, and where a shift of focus is possible. Then the conversation is:
It is ok we suck, how can we get better?
A lot of good soul quenching work happens after that. It feels good.
I hope it happens for you.
Much love.
Avinash.
|
Aiming to wow your employer with an outrageous amount of impact (or just shooting for a Much Higher salary)?
Upgrade to TMAI
Premium here - it is published 50x / year.
|
|
|
|
|
|
|
| 
