Brand media spend for startups

Measurement, formats and theory

Introduction

Brand media spend is a subject that causes a lot of debate in startupland. In my view, it’s actually quite straightforward. Brand media is powerful when done at the right time & in the right way.

With the recent story of Airbnb slashing all performance marketing spend and only spending on brand media - this subject can feel a bit disorientating. We're all bored of looking at Facebook Business Manager, but this isn’t the end of performance marketing for startups just yet.

It’s important to study brand marketing academics such as Les Binet, Mark Ritson and so on. But often their models don’t apply to startups.

What is brand media?

Firstly, to define brand media here - I'll be talking about any media spend that doesn’t focus on prompting the prospect to take a direct action. 

For most startups, this relates to "above the line" media like Out-Of-Home (OOH) and TV. But is also relevant to half brand, half performance channels such as Podcast Ads, TikTok and so on. You can even spend on Meta top-of-funnel ads as a brand media campaign.

Retail-first business

Weetabix hits you with multiple millions of pounds worth of brand advertising so that when you are in Aisle 3 on Sunday looking for an “allegedly healthy cereal” - you pick up their box.

There was a study that measured the brain activity for hardcore Apple fans and it lit up the same area as religious devotion (source).

These are great examples of where spending money on brand impressions leads to an increase in rate of sale in-store.

Generally, these brands have a lot of money - so they can get enough impressions to penetrate a large-scale customer segment. You need to have comprehensive retail distribution and big enough budgets to make the media work. If the campaign increases the rate-of-sale then happy days, if not then… macro factors? ;)

For challenger brands, other methods are prioritised to drive immediate results before they are ready to drop millions of pounds, for example: in-store pop-ups, point of sale physical advertising, giving away free products outside the shop and so on.

To get brand cut-through on smaller budgets, they often have to work PR harder by leveraging celebrities and influencers - or try to catch the right narrative wave, at the right time. Hype is their friend. Also, paid social and other performance channels are leveraged to increase brand awareness in the early adopter market (that’s often how ecommerce spend is justified for a retail-first business).

Digital-first businesses 

Let’s take a chronological example: a digital-first startup (eg. Consumer Fintech, DTC) wants to spend on brand media.

The first step here, in my view, is to spend on brand media and track it on a CAC basis. We will be getting to the traditional brand media measurement approach (Mark Ritson et al) once we’ve established our CAC.

This is because I want to:

i) Understand whether this activity is effective in getting people to purchase my product

ii) Understand what effect this less-efficient media has on the customer economics

iii) Help take stakeholders (eg. the board) on the journey, by having control over the numbers

The best case scenario for your brand media campaigns - the definition of success I go by - is roughly 3x the CACs you get from your core performance campaigns. That’s just a general rule of thumb for your first brand media campaigns.

Sometimes just explaining that to people is quite a sobering experience for them. I come across people who believe that their brand media campaigns will not significantly increase their short-term CAC. Furthermore, sometimes people believe that a small brand media spend will deliver results. Sadly, it can’t. It’s just a drop in the ocean. One caveat being challenger brands who use this as part of a wider PR campaign to influence retail buyers or journalists, but in that case arguably you only need one billboard and a camera (a common tactic, by the way).

I'm an attribution pluralist (jargon alert), which means I want to be looking at as many different views of performance as possible. If you gloss over the below list, the main takeaway is that there are a lot of powerful & different ways of looking at the performance data. 

For OOH and TV, you could be looking at I) post-checkout survey in isolation ii) an econometric model (like TV Squared) iii) an AI model that combines all data points iv) Media Mix Model (MMM) v) Brand uplift studies pre and post campaign vi) Spend incrementality studies.

And breathe. 

And why not look at various views? This is potentially the biggest spend of the company's life to date and biggest future cost. The best CAC analysis is done by triangulating multiple views. One attribution view to rule them all is arguably important for company management and comms. But behind the scenes, where the real work is done, it's okay if the marketing team looks at multiple attribution sources, as long as they are fluent in understanding their relative methodological strengths and weaknesses, in depth. 

This helps us understand how the profitability of our cohorts are likely to change and whether the advertising actually did the job of selling products. If you don’t do this, then you will have no idea how much working capital you’re tying up in the future.

If you are extending the payback period unknowingly, you may end up having to go earlier back to fundraising, cap in hand.

The profitability of acquired customer cohorts is the lifeblood of a startup. So we can’t just charge zealously into brand media because someone “knows it works”.

However, we know that brand media is not a short-term activity, it’s a rising tide that floats all boats. So we can give ourselves a lot of flex here.

We can start to analyse a fully-loaded, brand CAC over a longer time period - whilst keeping our core performance CAC still measured on a daily, weekly and monthly cadence. This allows us to focus on how this brand spend blends over a 6-9 month period, rather than prematurely calling the effect of the campaigns. 

A world beyond CAC

Once there is more confidence regarding the short and long term effect of brand media on our cohorts, and that intel is fed into forecasts - then we can move into a more traditional, brand media analysis.

How relaxing life is outside the mud-pit of CAC measurement. We’re like modern day Mad Men.

Not really. It’s always got to tie itself back to revenue. My favourite way of doing this has been popularised by Mark Ritson and others.

Awareness
Consideration
Trial
Repeat
Preference 

Measure each step of the funnel by running YouGov surveys with your target audience. Understand how penetrated you are in each part of the funnel. Because you can run a campaign before and after the activity, you can measure the increase in trialist / repeat customers and therefore the financial impact of the activity at the top of the funnel. Teams can start to build predictability of how the funnels will be impacted at different brand media spend levels. They can also run their data vs competitors, which allows you to see how the battle for share-of-voice is playing out in your target market.

All men (and women) are created equal, all media is not

We hold these truths to be self-evident. Bus, taxi, and billboards are for the purest brand awareness campaigns only. People only see these big formats for a flash and you have less space to communicate. The copy needs to be large. Hard to communicate a proposition.

If you’re a startup considering these formats, then you must be extremely late stage and playing a brand-awareness game that you know delivers ROI.

If it’s your first campaign, head to the soft landing that is “Tube Car Panels” (fun fact: called Tube Car Panels rather than Tube Carriage Panels as the creator of the London Underground was from NYC). This format has the highest dwell time (average around 13 minutes) therefore it’s great for landing more complex propositions and is known for having the highest direct response rate. You’ll often see brands leaning into this by using an offer code to amp up the direct response element. This will help build confidence with the board and other stakeholders in the media’s ability to drive purchases at a decent rate.

From there you can build to “6-sheets” (human-sized banners that follow you around the tube) and “16-sheets” (cross-track when waiting for tube).

In general, London TFL tends to perform much better than non-TFL media.

TV advertising

In terms of TV, it’s a bigger commitment as you need to spend on creative and most likely commit to a longer-term channel spend. Attribution and administration are more draining on the organisation than OOH. 

Not nearly enough space in this article to cover the basics. What I will say is that there is a big difference between:

i) Direct Response TV

Targeting the low volume, long tail channels across Sky and so on. Perhaps counterintuitively the less engaged someone is when watching a show (eg. cash in the attic on the background) the more likely they are to be dual-screening and might load up your website once they see the ad. Furthermore, this media is cheaper and therefore you’re more likely to see a short-term ROI.

ii) Brand TV

The high impact, ITV-big-show slots. People are very engaged and therefore less responsive on their second screen. Much more expensive media, therefore harder to see short-term ROI but has big brand equity potential.

iii) Sky AdSmart

Extremely targeted, great customer segmentation. You can target specific household demographics. However, it is so much more expensive that you better believe your conversion rate is phenomenal with this audience.

iv) Streaming ads

Last shoutout for ads that run on the online steaming platforms, such as channel4 and so on. Back in the day, I was quite excited when this came out because I thought the targeting would be unreal. But actually they don’t have that much data about you at all and I’ve never heard of any strong response rates. It feels like a channel that has the illusion of effectiveness just because you get click-through data and so on. I think it is very much pure brand media. If someone has a decent counter position, please message me.

If you want to follow the system I described previously of establishing your CAC for your brand media first - direct response TV would be the natural place to start. Build up confidence in the stakeholders, board members and so on - before you go big on a Brand TV play.

Lastly, there are quite a lot of TV media for equity deals around. The benefit is that you can save cash, the negative is that you’re locked into a particular channel and strategy. Not as much flexibility for testing. If you are giving equity away, it’s not free. I’d want to convert that back to £ in my own analysis to make sure I’m being honest with myself about the effect it is having on my cohorts.

Conclusion

To have any clear conversations about brand media, you need to define what stage of business we are talking about. The strategy, tactics and financial implications from early-stage to late-stage or from retail-first to digital-first vary greatly.

It’s frustrating when people apply late-stage or enterprise marketing principles without appreciating the specific situation many startups are in. It’s like bringing a paint brush to a knife fight.

Startups don’t often have the financial position to support a big, untargeted brand media campaign. Nor a proposition that is ready for the mass market.

That said, startups & growth people should study Les Binet, Mark Ritson and all the rest so that they are fluent themselves in understanding how the brand media game is played at scale.

Both groups have a lot to learn from each other. The best approach for startups is to have your eyes wide open about the pros and cons, rather than a zealous conviction either way.