How Much Of Your 2018 Social Media Spend Will You Waste? (again)

AdParlor Blog Post: How Much Of Your 2018 Social Media Spend Will You Waste? (again)

Everyone who has studied or worked in marketing will know the old phrase ‘I know that half my marketing budget is wasted, but unfortunately I don’t know which half’. In the old analog world, it was very hard to avoid waste – showing analog ads to people who would never buy your products, or showing ineffective ads to potential customers. Mass marketing – generating one ad for everyone, and pushing them through dumb ‘spray and pray’ media such as TV and newspapers – was understandably wasteful. However, surely digital marketing – with its precision targeting and ability to micro-segment – must have eliminated all waste by now?

Unfortunately not – because many of the old bad habits of mass marketing are being applied to digital marketing.


New Media Enables New Ways To Identify Waste

In 2017 AdParlor – a people-based marketing company – conducted over 50 ‘deep dive’ audits into the effectiveness of large advertiser social media and display advertising. The team had access to the ad accounts (to understand audience targeting, budget allocations, bids, results etc.), plus the ability to tag creatives to identify which elements were working and which not. The summary finding was that those 50 companies were wasting – or spending highly inefficiently – between 12-80% of their social media budgets. Most companies wasted 20-50% of their 2017 social media spend.

Some examples of waste:

  • Fixed Budget Allocation. For a large eCommerce company, their best performing ad-set for a mobile app install campaign only received 8% of spend. The other 92% was spent on significantly less effective ad sets, with much lower returns on investment (often negative returns). Had the budget been dynamically directed towards the winning ad set (with cost-per-install remaining constant) installs would have increased by 12x for the same media spend, because cost-per-install would have declined by 92%.
  • Ads Getting Stale. Across the board, most advertisers – on average – let ad creative go stale (i.e. they stopped generating results) for a month before changing them out. This is easy to monitor, and generating/ testing new ads is quick and cheap.
  • Overlapping Audiences. The AdParlor team almost always found overlapping ad-set audiences, often of 50-70%. Overlapping audiences causes mixed messaging to consumers and drives campaign inefficiencies. Reducing such overlaps typically increased campaign ROAS by 10-20%.
  • Limiting Ad Types. Many advertisers limit their use of the various ad types on offer – which is criminal on platforms such as Facebook, when so many ad formats are on offer, generating radically different returns. Some examples of opportunities identified during these audits included:
    • Dynamic Product Ads. Using Dynamic Product Ads, a jewelry retailer saw a 44% improvement in ROAS compared to other campaigns geared toward conversions.
    • Offer Ads. Using Offer Ads, a beauty retailer saw a 40% improvement in ROAS compared to identical Link Ads.
      Carousel Ads. For an online fashion company, Carousel Ads saw a 15% stronger CPA compared to Link Ads.
    • Retargeting. For a diverse eCommerce company, Retargeting audiences generated a 27% lower CPI (cost-per-install) compared to Interest and Lookalike audiences in driving mobile app installs.
  • Creative. The difference between good creative and bad is enormous in terms of performance, yet most marketers still ‘sign off’ on each element of creative copy by committee, rather than based on performance. Here are some examples of how creative has impacted performance:
    • Beauty Retailer: 28% higher ROAS for ads that stated "X dollars off" instead of "X percent off".
    • Apparel eCommerce: 25% higher ROAS when images featured no gifts compared to images featuring gifts.
    • Quick Service Restaurant: 92% Higher ROAS when ads featured main courses instead of appetizers.

Example of People-Based Creative at work: Using technology designed to mass produce tailored ads at scale, one beauty/ lifestyle omni-channel retailer created 130,000 social media ads during 2017. Of this number, only 5% of those ads contributed to goal conversions – the rest were quickly dropped or tweaked.


The Solution is People-Based Marketing

So how do you ensure minimal waste in this new world of precision targeting and auction-based bidding? The first is to set your entire marketing organization on a course towards people-based marketing. That means building the team, technology, capabilities and partnerships required to – one day – be able to efficiently deliver a different (tailored) ad to every single customer, based on detailed knowledge of that individual. You need to consign mass-marketing to the trash can, along with all people, practices and partners who are holding you back.

Tactically, somethings you can do to get your team started are as follows:

  • Expect a Return. Stop treating marketing as money that you spend with unknown returns. State ROAS goals, in order to benchmark performance of all campaigns. You may get these wrong, but the discipline of expecting a return from every dollar spent, and of benchmarking ROAS on every campaign, will start driving smarter measurement of return.
  • Segment More Ambitiously. Define 20+ tailored segments and customer journey triggers at any time. Ensure that each of these has a tailored ‘mini-campaign’ to maximize returns, including a LOT of testing (see below). Keep increasing the number of segments over time, and don’t let up until you are serving a different ad to every single consumer.
  • Create Many, Many More Ads. Ensure that you can mass create of tailored ads. That means at least 3-5 ads per segment, and more if required to beat ROAS expectations. Also try out multiple ad formats and the smart use of user-generated content (UGC). You will likely need a technology partner to mass-product tailored ads in the hundreds/ thousands.
  • Test, Optimize & Automate. Make it a norm for your team/ technology partner to continuously test different audiences, creative, ad types, bids, landing pages etc. Also utilize automation/ machine learning for this purpose.
  • Apply Beyond Social. Use your learnings from social media to 1) generate massive consumer insights to improve strategy, and 2) apply to other media, e.g. display campaigns, to improve the ROAS from those too.

This is a LOT of work, and can’t be done with your traditional manpower levels, or without the help of powerful technology, but if your media spend is meaningful, the ROI improvements on your media spend totally justify these additional costs.


Next Steps

Interested in working with our Team? Need expertise in Media Buying, Creative Services and Auditing your Campaigns for 2018? Chat with us below by clicking here.