Welcome to Marketing BS, where I share a weekly article dismantling a little piece of the Marketing-Industrial Complex — and sometimes I offer simple ideas that actually work.
If you enjoy this article, I invite you to subscribe to Marketing BS — the weekly newsletters feature bonus content, including follow-ups from the previous week, commentary on topical marketing news, and information about unlisted career opportunities.
Thanks for reading and keep it simple,
Models versus experiments
For the last six weeks, I have highlighted ways to apply marketing concepts to pandemic responses. Today, I pivot back to a more “pure marketing” question that I’ve been repeatedly asked by Marketing BS readers and the portfolio companies I work with: “How should we think about marketing during this crisis?”
In last week’s letter, I wrote about the limitations with models and the value of experiments. I received lots of feedback from readers questioning the difference between a model and an experiment.
Here’s an email that articulated many of the relevant ideas:
Aren’t models and experiments just two parts of what used to be called the ‘scientific’ method. Experimenting without a hypothesis is really just a demonstration — and while you can sometimes gain insight from this type of inductive activity, it doesn’t lead to explanative deductions that can be knowingly repeated. Modeling without experimentation is just a theory. Both are needed.
Those are some great points, and I recognize the blurriness between models and experiments.
So, I would like to clarify my arguments from last week’s post. When I wrote about the limitations of “models,” I was specifically referring to the current trend of trying to take complicated analyses and force the results into a curve. Visual representation of these models — regularly shown at press conferences or included in public health materials — are incredibly powerful; they influence governments to implement restrictions and they convince the public to stay at home. And yet, despite the importance of these models, there are no clear explanations of HOW the results were determined or the possible limitations of their analysis. Moreover, the methodological details used to create these models usually sit in a “black box” that is not accessible to the public (or sometimes, not even people on the research teams).
I would contrast that definition of a model to one where people try to make informed decisions based on data. Not sure which email format improves sales? Try an A/B test. After a few rounds of tests, your analysis can reveal clear trends (e.g., emails with emoticons in the subject line boost open rates). From there, you can develop a “model” to help predict future outcomes.
Given the importance of models to so many fields, I expect to return to this subject in future newsletters.
People are buying soap!
An April 17 article in The Drum reported on Procter & Gamble’s recent sales activity:
P&G today (17 April) reported an organic sales increase of 5% year-on-year as the coronavirus boosted consumer demand for products in its healthcare, fabricare and homecare categories.
The CPG giant, which produces the likes of toilet paper brand Charmin and cold relief medicine Vicks, witnessed US sales rise by 10% this quarter. The bump successfully offset the corresponding 8% decline in the Chinese market.
Despite the lift circumstantial demand — as well as “concentration” of the number of products shipping to retailers — chief financial officer Jon Moeller confirmed the company would be putting its foot down on media spend, rather than taking it off in an effort to bank cash. [Emphasis mine]
During every recession, companies experience a shift in the distribution of consumer spending. Different product segments experience varying types of impact, but there are a couple of common trends:
A delay in purchases for large durable goods. In periods of uncertainty, few people buy new cars or dishwashers.
A continuation in purchases for consumable products. No matter the economic situation, people still buy toilet paper.
The COVID-19 crisis appears to have affected the distribution of consumer spending in particularly extreme ways. The New York Times analyzed credit and debit card data to research changes in spending since January. The results are striking:
Reports about P&G’s recent bump in sales should not surprise anyone who has walked through a grocery store recently; you can see that customers continue to purchase paper towels, toilet paper, and cleaning products.
Meanwhile, check out the trajectory for the bottom line on the NY Times chart. Without question, COVID has devastated the travel industry.
So, if you are responsible for marketing decisions at a company — let’s assume for an industry somewhere between “toilet paper” and “in the toilet” — how should you allocate and execute your marketing spend? And how, if at all, should you adjust your brand messaging?
Everything is different, but nothing changes
In 2015, social psychologist Philip Tetlock’s Superforecasting: The Art and Science of Prediction (co-written with journalist Dan Gardner) outlined a methodology for predicting future events.
Suppose you wanted to calculate the likelihood that an event will occur, such as “an Israeli election in the next eight months.” You could begin by assessing the current political gridlock in that nation’s electoral system.
More effective forecasters, on the other hand, use a process referred to as the “outside view.” Instead of scrutinizing today’s polling numbers, they begin with questions like, “how often does Israel hold elections?” and “what is the variability of the length between elections?” Only AFTER acquiring a deep understanding of “normal” patterns will the researchers adjust their perspective based on present-day factors.
Their general advice for improving the accuracy of your predictions?Information about “what is happening right now” should be underweighted, contrary to what your instincts might suggest.
How can we apply this concept to projections for a post-COVID world? You’ve probably read dozens of think-pieces that predict a future where EVERYTHING will have changed.
According to the principles of “superforecasting,” though, we should actually expect FEWER — and less substantial — differences than most people imagine. We should begin with the default assumption that the pandemic and accompanying lockdowns will not impact society in pervasive ways. Starting with that “base rate” assumption, we can dive into the details to see if, or where, we might expect changes in behavior or policy.
To begin, we can review shifts in customer activity from BEFORE the pandemic. Recessions tend to accelerate trends that were already in motion. For instance, the Great Depression of 1929 delayed the penetration of durable household goods like telephones. In present day, the stay-at-home guidelines will likely boost the already growing adoption of e-commerce and remote work.
But some of the bolder claims about the future — like a massive reduction in people’s desire for restaurants, concerts, and travel — seem far more speculative and less likely to occur.
How can you apply these ideas to your own company? For most of you, it’s reasonable to believe that once we move past the coronavirus situation, your business will resemble its position from January 2020.
On April 21, Netflix conducted a quarterly earnings call. CEO Reed Hastings announced better-than-expected subscriber growth in Q1 — Netflix gained 15.77 million new subscribers, far more than the company’s original projections of 7 million. But Hastings also cautioned that many of these new subscriptions might just have been “accelerated.” In other words, Netflix picked up customers who would have likely joined the streaming service in Q2, Q3, and Q4; those people signed up earlier than expected because of the sudden and atypical need to stay at home.
The accelerated subscriber growth in Q1 might suggest some disappointing numbers for Netflix in the coming quarters; as Hastings stated, “It's essentially a pull forward of the rest of the year.”
But, the CEO also stressed the continuity of their long-term projections: “will Internet entertainment be more and more important over the next 5 years? Nothing has changed in that.”
The pandemic accelerated Netflix’s business — but the company was already trending upward. Nothing has fundamentally changed.
Show me the dollars
One area that we CAN expect change: marketing strategies. In one particularly extreme example, travel giant Expedia expects to cut its 2020 advertising budget by a minimum of 80%.
Before thinking about how much your company should spend on marketing, take a step back and look at the bigger impact of COVID-19 on marketing costs and consumer activity.
The good news: your marketing dollars will go further. The cost for an impression (CPMs) on Facebook dropped 35–50% from early March. For Google, the effect on costs per click (CPCs) varies by product category; prices for many segments are down more than 30%. Collecting reliable information about the cost of television ads is more difficult, but some estimates project a decline of 40% in Q2 advertising dollars for TV — despite the fact that total viewing time is soaring.
The bad news: your customers may not be responsive to making purchases at the moment.
The REALLY bad news: your business may not have enough cash to survive until the time that customers return to their previous buying behavior.
With those ideas in mind, your company’s number one goal should be clear: focus on your ability to survive until the end of the crisis. Bankrupt companies no longer care about the level of their brand awareness.
Unfortunately, no one knows the timeline for pandemic-related restrictions and economic turmoil. Optimistic projections show a (mostly full) recovery by Q3, but other analysts expect a return in 2021 is more likely. If your business cannot survive on life support through 2021, you should look very carefully at where your money is going out the door. (And Bill Gates believes it could be FIVE YEARS until life goes back to “normal.”)
For companies on a safe footing, you can look for the smartest ways to spend your marketing dollars.
Although every media article seems to includes the phrase “unprecedented times,” savvy marketers should actually review “precedent” — the research on what worked during previous downturns. In the midst of the 2008 recession, marketing expert Byron Sharp published a comprehensive summary of the ways that customers behave during periods of economic uncertainty. I highly recommend checking out his full report; the insights are still valuable today (maybe even more so).
Here are some key ideas from Sharp’s article:
One of the big lessons we’ve learnt from behavioural data (category buying, brand buying, and media consumption) is how habitual consumers are. Underneath the day-to-day randomness lie rather consistent patterns of behaviour. This suggests that even if consumers want to, they will find it difficult to change their behaviour.
What this suggests is that in ‘belt tightening’, consumers will find it easier to make changes in non-regular behaviours than to substantially modify their day-to-day lives.
The 1970s saw probably the most extreme recession since WW2, with an awful combination of high oil prices, high inflation, high interest rates, high unemployment, and negative economic growth (4). Yet Nielsen records over this period do not show a contraction in purchases of consumer packaged goods. [Emphasis mine]
On the subject of pricing, Sharp suggests that relative prices matter. If your competitors have not cut their prices, then you do not need to cut yours either. In the event that your rivals do raise their prices, you have two major options: (1) keep your prices the same and increase market share, or (2) raise your prices to a similar level and earn additional margin. Consistent prices might shift your market share higher or lower, but you will only see the impact over a lengthy period of time.
Moral of the story for your marketing plans: don’t think that you can reduce prices to steal share, and then — post-coronavirus — increase your prices AND hold onto the new market share you gained.
Sharp also offers some perceptive comments about marketing spend during times of crisis:
There are a number of studies that purport to show that firms that spend more during a recession do particularly well. However, these have been criticised as being funded by vested interests (e.g. media companies), using cross-sectional data, and inappropriately attributing causality to ad spend when something else (e.g. an already growing brand or category) may be the real reason for rising ad spend and sales.
However, a 2003 MSI report based on a longitudinal analysis shows that firms who increased advertising spend during a recession did slightly better than firms who increased advertising spend during other times. Equally importantly it showed no effect of cutting advertising during the recession, perhaps because this was the norm (i.e. most competitors cut advertising). …
[W]hat we know about normal periods is that share of voice (SoV) matters.
Recessions offer the opportunity to substantially increase SoV if competitors cut their ad spend and media prices drop. [Emphasis mine]
Research confirms that if you can afford to increase marketing spend during a recession, you will see a greater impact than during other times. I believe this theory holds true for two reasons, both mentioned above:
Media prices are cheaper during a recession, so you get more bang for your buck. Keep in mind that the short-term impact of your campaigns will be worse than normal. If you buy an ad with the goal of getting someone to buy your product, the conversion from click-to-purchase will likely be lower than it was two months ago (because people are being more careful with their money). As such, your measured impact — revenue per click — will be lower. Over the long term, though, your ads will (probably) increase awareness and consideration of your product with similar “impact per impression” as they did before the crisis.
Competitors are likely cutting back, if only so they can stay alive. When your “share of voice” increases, you can win disproportionately to the number of impressions you obtain. As your competitors move out of awareness, you can win consideration by default.
What should your ads look like?
During the pandemic, you should consider altering the aesthetic elements of your brand messaging. Your commercials should start with sparse and somber piano notes, before revealing black and white images of empty streets and venues. With the opening tone established, introduce a voiceover that shares inspirational messages like “even though we are physically apart, we remain together in spirit.” Make sure to mention people, people, people, and “uncertain times.” Feature emotional footage of children and grandparents, first responders, and the brave employees of your company. Remind people about the historic presence of your brand in their day-to-day lives and family traditions.
Emphasize that your product — no matter what it is — can help bring people TOGETHER. Conclude the commercial with uplifting musical compositions (think Olympic Games) and a focus on people clapping from their balconies to support health care workers. (Insert emoji for sarcasm…)
If you’re stuck on executing those ideas, you can watch this mash-up video of many new ad campaigns that recently aired:
Keep it simple and stay safe,
A few weeks ago, I shared InterExec’s impressions of COVID-19’s impact on executive recruitment. They recently updated their perspectives with some specific quantitative data:
APAC: The market is still slow but certainly picking up, particularly in Hong Kong. Change/transformation/problem solving are in demand.
Middle East: Demand is generally seen to be about 75% down with a massive slowdown in Construction and Chemicals.
US: Circa 75% down and in particular Oil and Gas is not coming back any time soon. Expected to be worse than 2014.
Western Europe: 50%-70% down. Very few assignments are being put on hold, but few new ones now coming in. FMCG, Digital, Healthcare, Logistics and Retail are still strong, Industrials and naturally Hospitality, Leisure and Airlines are particularly hard hit.
Eastern Europe: 90% down, generally with urgent hires only.
UK: 70% down. With only business critical new assignments. Turnaround, Transition and Project Management will be best. Construction, Infrastructure and Power are particularly hard hit. In Leisure and Retail, some recruiters are exiting the market. There is some optimism for June.
If you enjoyed this article, I invite you to subscribe to Marketing BS — the weekly newsletters feature bonus content, including follow-ups from the previous week, commentary on topical marketing news, and information about unlisted career opportunities.
Edward Nevraumont is a Senior Advisor with Warburg Pincus. The former CMO of General Assembly and A Place for Mom, Edward previously worked at Expedia and McKinsey & Company. For more information, including details about his latest book, check out Marketing BS.