Instagram Beats Google+ by a KO: Social Ads Drive 2,885% Higher Engagement in Early Test
In the last few weeks, both Instagram and Google+ have announced their ad platforms and opened them as “beta” programs to select brands. Amongst those brands are two of the biggest names in the automotive industry: Toyota & Lexus. ToyotaUSA participated in Google’s +Post pilot campaign, which was featured as the flagship campaign when Google announced its new +Post platform. Lexus, on the other hand, participated in the highly covered Instagram sponsored post launch alongside brands like Michael Kors (which got most of the media attention).
Both brands have run paid ads on the respective networks in the last few weeks. From an initial analysis of the pilot campaigns, we found that the Lexus campaign, ran on Instagram, outperformed ToyotaUSA’s campaign on Google+ on almost every single aspect, including total engagement, followers growth and most importantly, post-campaign organic lift.
While this is not a classic “apples-to-apples” comparison, the lessons from this analysis are interesting nevertheless and can help you decide what network you should spend your money on.
ToyotaUSA on G+
We focus on two Google+ ads ran by ToyotaUSA on 11/27/2013 and on 12/2/2013. On average, the two sponsored posts saw a 228% increase in total engagement over the running average of the last 40 days. More specifically, since November 1st, ToyotaUSA has had an average of 48.5 engagements per post (not including the +Posts), while the two sponsored posts saw an average engagement of 159.5, a 228% increase in total average engagement.
We calculate Engagement as the sum of +1’s, comments and reshares per post on G+. ToyotaUSA saw a lift in all three metrics, but most impressively were the reshares. On average, ToyotaUSA generated 2.3 re-shares per organic (non-paid) post, but its sponsored posts saw an astonishing lift of 3,402% to an average of 80.5 reshares per sponsored post.
However, these results had no permanent effect on ToyotaUSA’s overall engagement. Engagement levels (across all three metrics) returned to “normal” levels (within an acceptable margin of error) and saw no additional lift outside of the sponsored posts.
ToyotaUSA also received several comments on its sponsored posts. While the number of comments was not overwhelming high (average 5.5), all 11 comments, with no exceptions, were positive.
In terms of circler growth (the number of users adding the brand to their circle), we saw no impact on ToyotaUSA’s circled growth rate. In fact, ToyotaUSA’s circled growth rate has seen almost no major spikes or dips, averaging a steady 0.36% daily growth since November 1st, 2013. On the days the +Posts ran, the average circler growth was 0.34%, within less than 1 standard deviation of the mean.
Lexus on Instagram
For this analysis, we focused on one sponsored post Lexus promoted on Instagram on 11/22/2013. On average, Lexus’s engagement levels on Instagram are significantly higher than ToyotaUSA’s engagement levels on G+, but since we’re not comparing overall engagement levels, just the impact of ads on engagement, we used the running averages as the baseline and looked at the lift.
Lexus’s sponsored post saw an incredible 6,807% increase in total engagement. More specifically, since November 1st, 2013, Lexus has had an average of 3,153 total engagements per post – an impressive level by itself- but its sponsored post saw an almost-unbelievable lift in engagements to a total of 217, 767.
On Instagram, we calculate engagement as the sum of comments and likes, with sharing on Facebook and Twitter (to set a benchmark against reshares on Google+). The vast majority (96%) of engagement on Lexus’s Instagram posts occur on Instagram, with only 4% of engagement occurs on Facebook and the rest (less than .001%) on Twitter. This distribution stayed the same before, during and after the campaign.
Since November 1st, 2013, Lexus saw an average of 3,022 total engagements per post on Instagram, with an average of 2,959 comments and 63 likes (we took out the ad campaign stats from these averages as they were more than 100 standard deviations from the mean). The sponsored post we analyzed saw an incredible lift in all metrics across the board. Total Likes jumped by 6,713%, from 2,959 to 201,584; Comments saw the biggest increase of 10,769%, from an average of 63 to 6,842; and shares on Facebook and Twitter moved from an average of .6 average to just 5, still recording a 726% increase.
However, these numbers tell only part of the story. When we glanced over the comments (given the phenomenal increase in the total number of comments), we noticed that a lot of them were negative; users reacting to the sudden appearance of ads in their feeds. We extracted a random sample set of the comments (n=308) and found that 64% of the comments were negative and only 9% were positive (27% were either neutral or undecided). The negative comments were directed at both Instagram (and Facebook) and Lexus as users took to the comments section to show their astonishment and disapproval of ads showing up in their Instagram feeds.
But while the number of negative comments might be a point of concern, the most amazing (and desirable to any advertiser) result of this campaign comes up when comparing Lexus’s average engagement levels before and after the campaign.
Overall, average total engagement per post grew by 60% for posts published after the ad campaign. Broken down by engagement type, total Likes grew by 61% (from 2,302 per post before the ad campaign to 3,705 per post after the ad campaign); total comments grew by 31% (from 53 to 70); and total Shares more than doubled from 0.4 to 0.83. Unlike the effect of the G+ campaign on ToyotaUSA’s ongoing engagements, Lexus campaign on Instagram drove a real lift in Lexus account’s organic (non-paid) engagement.
As far as audience growth, Lexus benefited greatly from the ad campaigns. On the days following an ad campaign, Lexus grew its followers by an average of 9.3%, adding an average of 3,862 followers.
Social Ads Showdown: Early Test Results
Did Instagram really do so much better?
From the numbers above, the answer should be an obvious “Yes,” but the huge differences in performance warrants a closer examination of the reasons for why Instagram ads performed so much better than Google+. This could also be a one-off comparison and not standard.
My immediate reaction to these numbers was that the audience on Instagram is just much more engaged (and larger) than the audience on G+. But while this might be the case, it doesn’t explain the huge difference in percent change – the lift in engagement – that these two campaigns saw. We ran a quick analysis on the engagement as percent of audience to gauge on the baseline engagement organic-posts get.
ToyotaUSA has an average of .03% engagement as percent of followers (194K users have ToyotaUSA in their circles) while Lexus has an average of 6% engagement as percent of followers (over 54K followers on Instagram). This by itself could explain some of the differences in engagement, but Lexus’s ad drove that number to over 400% (!) while ToyotaUSA increased it only to .08%.
A simpler explanation could be the number of impressions that each ad received during the campaign was very different. Perhaps the Lexus ad received much more impressions to start the campaign, while the ToyotaUSA ad was limited by targeting and budget (as well as by Google)? Since we don’t have the number of impressions served per ad (proprietary data to the ad platforms, and more than likely a factor of the brands bid levels and campaign budgets), we can only assume that this was part, if not the main reason for these differences.
Another explanation is the viral nature of social networks. More specifically, Lexus ad’s level of engagement was the cause for more engagement. As the Lexus ad took off and more people engaged with the ad, more users got exposed to it through their personal networks engagement, propelling the ad to a viral level that contributed to the ongoing engagement and drove even more engagement. ToyotaUSA’s ad did not reach that virality level and was subject only to the brand’s paid effort to push the ad to more users.
Additionally, there is a difference in the way the two networks enable sharing (and virality) which might have contributed to the huge differences in total engagements. Google+ allows users to +1, comment AND share (re-share) a post while Instagram only allows users to Like or comment. Re-sharing on Google+ means that the post is no longer directly associated with the original author and so are all the engagement metrics that will happen as a result of the re-share. So while brands can benefit from the extra exposure re-shares give them, they have no way of capturing that engagement and “keeping it in house.” On the other hand, Instagram, by not allowing “re-shares,” allowing brands to capture and leverage the entire engagement a post generated.
Lastly, Lexus’s ad benefited from the freshness and controversy of Instagram ads. As users were exposed to this new ad platform, they were caught by surprise as Lexus’s ad showed up in their feed. An overwhelming number expressed their disapproval on the comments, but as one of the users put it “While there are barely 7,000 of you complaining about this ad, there are almost a quarter of ONE MILLION people who liked this pic. Let that sink in.”
Note: as a marketer, I would be remise if I don’t credit some of the difference in performance to the marketers behind the campaigns. However, both campaigns were creative, smart and professional. Some would argue that ToyotaUSA’s campaign was actually better (better targeting, better ad copy and creative), but as a data marketer, my ultimate test is the numbers.
This is just the beginning. As preliminary as this analysis might be, it indicates how powerful the ad platforms are in driving increased engagement. Both networks only started rolling out their ad products and both are already seeing success.
It’s hard to argue that ToyotaUSA’s campaign failed. It drove real lift in engagement and unlike the Instagram campaign, it drove strong positive sentiment from its users. In addition, the number of re-shares were extremely positive, even though this analysis doesn’t cover the full extent of that impact.
Context (Target + Relevant Content) = Positive Engagement. While ToyotaUSA’s ad performance pales in comparison to that of Lexus on Instagram, it’s clear that their ad was much better targeted and had better content for its target audience. If you want to drive positive engagement, make sure you create unique, highly relevant content for your target audience.
Any publicity is good publicity? Lexus used the freshness and controversy of Instagram ad platform to drive record engagement on its account. It added over 14K followers to its account and saw a 60% lift in average engagement per organic post in the days following the campaign. These are highly desirable stats for any advertisers and they might be worth the negative sentiment that was directed mostly at Instagram and not at Lexus. Brands will have to decide on what they are willing to sacrifice in sentiment in order to see these types of returns.