Authors

Background

Why co-advertise?

Co-advertising occurs where more than one brand is included in an advertisement. Brand combinations in advertising can take many forms (see Appendix A). There are three commonly cited advantages of co-advertising:

The brand buyer advertising attention bias – Co-advertising can take advantage of the finding that brand buyers are more likely to notice advertising for brands they buy (Harrison, 2013; Vaughan et al., 2016). Having a second brand with a different buyer base expands the number of users who should be pre-disposed to notice the advertising. For example, Monopoly releasing a Game of Thrones branded board game to reach the Game of Thrones buyer/viewer base in addition to those who play Monopoly.

Halo effects – If a brand has a desirable quality, then partnering with that brand might enable the desired qualities to ‘rub off’ on our brand. For example, Pedigree partnering with pet-adoption agencies might help the brand be perceived as more caring about animal welfare.

Messaging effectiveness – A second brand can act as an additional creative device to help more effectively tell a story/communicate a message. For example, Pepsi using comparative advertising to showcase taste superiority over Coke, or Sephora advertising with Fenty to reinforce a message of catering to all skin colours.

In this research, we examine whether co-advertising delivers on two key metrics for advertising effectiveness:

• Advertising memory – do more people remember the (co-)advertising (in an easy ad recognition task)?
• Brand recall – among those who remember the ad, how often do both brands get recalled?

We see ad and brand memory as being important precursors to other co-advertising advantages, such as the aforementioned halo and messaging effects.

Research method

Our research involved 189 studies into the effects of co-advertising. These experiments followed a similar protocol, as split sample online experiments, with digital ads as the stimuli. In total over 17,200 respondents from the US, UK and Australia participated in the studies. Respondents were recruited via professional respondent panels.

The stimuli involved mock up pro-forma webpages featuring real news stories and mock advertising at the top and on the side of the webpage. Respondents were exposed to a series of webpages and asked to simply read the new stories within them. For each story/page, an advertisement was manipulated whereby respondents received either a dual branded execution or a single branded execution for a given brand. After a distractor task, respondents were then asked to identify if they had seen any advertising executions, with debranded visual creative used as the stimuli. Those that remembered the advertising were asked to recall any brands being advertised.

For the portfolio advertising, those that recalled the advertising were also asked if they remembered any products, so we could capture variants remembered.

Table 1: Outline of studies

Results

Brand usage and memory for advertising

A well-established empirical pattern is that users of a brand are more likely to remember advertising for that brand than non-users (e.g., Vaughan et al, 2016). For dual branded ads, we had three levels of usage: non-user of both brands, user of only one brand, and user of both brands. We checked all levels to see if the empirical generalisation holds for co-advertising, and if deeper usage is linked to greater memory for advertising.

The results (see Figure 1) confirm the hierarchy of memory, with non-users having the lowest ad recognition, single brand users in the middle, and dual brand users having the highest ad recognition. All viewers were exposed to the same co-advertising treatment, so this suggests the difference is in how the advertising is encoded in memory. That this effect is also evident for comparative advertising, where both brands are from the same category, reveals it is about usage of the brands, not about the creative of the advertisement.

In portfolio advertising, the difference between using one and both brands is lower because it is really only one brand being advertised (i.e., all variant users are by default brand users).

Figure 1: Ad recognition across brand usership for dual brand advertising

However, for brand recognition, the results show no difference (see Figure 2). Overall, advertising with two brands has the same ad recognition as advertising with one brand.

Figure 2: Ad recognition for single versus dual brand advertising

A deeper dive into the results shows that only 4% of the dual brand executions have 5pp or more ad recognition. The vast majority (81%) of dual branded advertisements have between +/-4pp viewer memory of their single branded counterpart, while 15% actually had over 5pp lower ad recognition.

Why no uplift in dual brand ad recognition?

It seems counter-intuitive that brand users are more able to remember advertising, but more brand users does not lead to more people remembering advertising. This occurs because while co-advertising gains more memorability from among users of both brands, it loses memorability among those who use only one of the two brands. The reason for this is likely that the presence of the second brand causes cue-overload, whereby users of only one brand get distracted by the second brand, and can fail to process that it was the brand they use in the advertising.

We see this when we compare the ad recognition scores among user of both brands in a dual brand context (overall 29%) and users of only one brand (overall 20%), with those who use only one brand in a single brand context (overall 23%). While ad recognition for users of both brands in dual branded executions is on average 9pp higher than users of only one brand, the score among users of one brand are 3pp lower than that for users exposed to the single branded execution. There is a systematic effect of lower recall on single brand users when a second brand is present.

And because in most cases, more people use only one rather than both brands, any gains from the buyers that use both brands is negated by the losses of the many more buyers that only use one of the brands.

Summary: Memory for advertising exposure

These results tell us that it is rare to get an uplift in advertising memory from advertising more than one brand, and that advertising with another brand to attract the attention of that other brand’s customer base is unlikely to pay off without careful selection of the partner.

If increased ad memorability is the goal, then pay attention to the overlap in the user bases of brands, that is the absolute number of how many use both, rather than the cumulative penetration of brand users, which is the sum of both dual and single brand users.

This suggests that co-advertising is unlikely to be an effective strategy for any small brand, nor is it of value for a bigger brand to partner with a smaller brand. Small brands have a small user base and so are already fighting upstream for attention, adding another brand to the advertisement just makes that battle even harder.

Brand recall

Our next test is about how often both brands are recalled post exposure. It is important to remember our test is a ‘best case’ scenario as we capture recall shortly after exposure in an experimental setting. In real life, any heightened brand recall needs to last longer than a 20-minute distractor task.

Our results show that among those who remembered seeing the ad, very few (5%) recall both advertised brands. Across the different co-advertising combinations, the highest this reaches is 19% for Comparative advertising. The lowest is 2% for Sports Events Sponsorship. In fact, the most common outcome was no brand recalled (50%), followed by the recall of one brand only (39%)¹.

When only one brand was recalled, this tended to be the brand congruent with the ad’s creative context. Ownership of the context occurs when the creative is linked to a given brand or its product category. For example, both Figure 3 and Figure 4 are advertisements promoting a partnership between Starbucks and Spotify. The first ad shows Starbucks owning the creative context by featuring a cafe backdrop and the Starbucks cups, whereas the second ad is owned by Spotify given music artists are shown.

Figure 3: Creative context owned by Starbucks

Figure 4: Creative context owned by Spotify

In our research, we created different co-advertising treatments to allow for each brand to own the context. Table 2 shows that brands that owned the creative context are recalled 4 times over those that did not.

Table 2: The impact of brand category/creative congruency on brand recall

Is recall related to brand penetration?

We tested whether bigger brands get recalled more, and if larger cumulative penetrations lead to a greater chance of both brands being recalled. The results show that while bigger brands that owned the creative context have the highest chance of being recalled, having a higher cumulative penetration did not enhance the recall of both brands.

This suggests that while two big brands might lead to greater ad recognition, it won’t lead to greater brand memory. It is still a battle for retrieval with the brand that owns the context more likely to be retrieved.

Summary: Brand recall

Partners on paper turn into competitors in memory when brands advertise together. It’s rare for both brands to be retrieved, and this is probably because they lack associative links in category buyer memory. It may be that over time, and with repeated advertising exposure, the necessary memory links are formed. However, given that it is challenging enough to reach and get the attention of a category buyer once, having an advertising strategy that relies on multiple exposures over time seems a risky and costly proposition.

The special case of Portfolio co-advertising

Portfolio advertising is a special case as both the core brand and variant share the same brand name, yet the variant has an added challenge of needing buyers to remember the variant as well as the core brand name. We see this challenge play out in the brand recall scores across different scenarios.

First, even when not present, the core brand dominates recall (see Figure 5), getting 2-3 times more recall than the variant advertised.

Figure 5: Brand recall for single brand advertising

When the core brand is added to the advertisement, unsurprisingly the core brand’s recall goes up, particularly if the creative context of the ad is more aligned with the core brand (Figure 6). However, variant recall goes down when the context of the ad focusses on the core brand. If the goal is to maximise variant recall, then advertising the variant alone, or alongside the core but with the context highlighting the variant, provide the best outcomes.

Figure 6: Brand recall for dual brand portfolio advertising

If the goal is to maximise portfolio recall (where either the core brand, variant, or both, are recalled), and you are less concerned about which brand is remembered, then any type of portfolio dual branding will achieve this (portfolio recall scores of around 90%).

Table 3: Comparison of Total Portfolio brand recall across advertising treatments

Summary: Portfolio advertising

The core brand will dominate recall regardless of whether it advertises alone or with a variant. So if the aim is to maximise variant recall, then advertise the variant without the core brand, or ensure it owns the context when it is paired with the core brand. For example, if Uber wants to advertise its pet service ‘Uber pets’, it needs to show the Uber brand, but focus the advertisement on transporting pets and highlighting the name ‘Uber pets’ so buyers know what button to press on their app. Both Uber and Uber Pet brands are present, but the context is about the Pet aspect of Uber. This means while many viewers will still just remember Uber, there is the maximum chance that any viewer will also remember Uber Pet. And those that just remember Uber, sill have refreshed that brand in memory, and so increased the chance of being bought at a later point in time.

If the goal is to maximise total portfolio recall, then advertise the core brand alongside the variant but it does not matter who owns the creative context.

Overall summary and key actions

On the surface, co-advertising seems like a win-win, providing a greater audience and shared memories. Unfortunately the evidence is not so rosy. We find typically:

• Ad memorability benefits are only limited to a small range of circumstances where there is high overlap in the user base of both brands.
• Irrespective of the brand pairs, they become competitors in memory for retrieval and the brand that owns the creative context is the most likely ‘winner’. Ultimately, the best course of action to enhance both brand’s recall is to give each brand their own advertisements to shine in.

Therefore, we recommend that marketers catalogue the co-advertising taking place and see if this can be reduced (see Appendix A for a list of co-advertising combinations).

If co-advertising is unavoidable:

• Advertise with a brand you have high brand usership overlap with.
• Own the context by making the advertisement look like your brand (Distinctive Assets help here) and/or its category.

Portfolio advertising is the exception here, the total portfolio is best served by co-advertising a core with a variant, but the variant is given the best chance of success if the context of the advertisement is congruent with its point of variation, whether that be a flavour, scent, pack size or price point.

______________________________________________________________________________

¹ A further 6% recall incorrect brand(s).

______________________________________________________________________________

Appendix A: Types of Brand combinations in advertising

______________________________________________________________________________