Do you forecast social media results?

Forecasting results is part of every campaign planning. We see “estimated results” in proposals and media plans all the time. During the planning phase of digital campaigns, many clients and brand owners consider what social media metrics are important to them and set them as Key Performance Indicators (KPIs) targets. They may also compare the KPIs proposed by vendors as part of the pitch selection process.

This is such a normalised aspect of campaign planning that it’s worth questioning how important are estimates. These questions are also relevant for new marketers who may be wondering how estimates are calculated and if their estimates are accurate enough. (But as already shown in the title,) I’d argue that estimated social media results are ultimately futile when planning for your campaigns. Hopefully, by the end of this article, you’ll see that estimates don’t give you an accurate picture, don’t set a constructive approach to campaigns, and restricts brands to think inside the box – so we really should not give estimates a hoot.

The reason why most businesses want to know estimated results

Estimates are a convenient way to account for marketing expenses (and get their marketing budgets approved); Estimated KPIs allow them to see a fair exchange, trading their budget for results. There’s something about metrics that seem so tangible, and therefore achieving (or aiming to achieve) something tangible appears worthwhile.

Unfortunately, there isn’t a reliable or standard way to come up with estimates.

Estimations or Guess-timation?

The 3 common methods marketers use to come up with estimates are:

  1. Cross-industry estimation: They look at broadly similar types of campaigns regardless of the industry, product or services to derive an estimate. For example, what an average website link click would cost regardless of what the brand, destination or platform.
  2. Intra-industry estimation: They look at results from competing brands, products or services to derive an estimate. So for example, toothpaste marketers may look at other toothpaste brands to derive an average result.
  3. Past campaigns estimation: They look at the brand’s past campaigns, especially if the content, creatives and angles are similar, to derive an estimate. If the creatives, angles, brand, budget and target audiences are unchanged, so should the outcome – but as you’d know this is rarely the case. The more uncontrolled or unknown variables there are, the less accurate a prediction would be.

Even if marketers have had tons of experience working with similar campaigns, their experience may be outdated or not easily translated to another brand. So these 3 methods would serve as a shoddy benchmark at best, and not an accurate estimation.

In most situations, marketers do not have data to refer to at all. In this case, they may opt for an,

  • “Optimal results” approach: Basing estimates on either an acceptable response (from the top of my head: a 2-5% click-through rate) or whatever the brand owners would be happy with.

For marketers, this… satisficing method… is just performative and expedites the approval of a marketing budget, which despite having little predictive quality, has a high practical value – so use it sparingly.

Marketers err on the conservative side

Whichever methods were used to derive an estimate, you can be sure that marketers would always err on the conservative side. They’d rather over-promise and under-deliver than vice versa because under-promising gives them leeway to go above and beyond the results they’ve committed to. Over-promising may result in termination if they don’t meet these unrealistic targets. So achieving (or even exceeding) the estimated social media KPIs does not equate to a successful marketing campaign.

Goodhart’s Law says it all

“When a measure becomes a target it ceases to be a good measure.”

Targets change behaviours. If a measure becomes a target, Marketers would be incentivised to game the system by optimizing for these measures, rather than what’s good for the brand or customers. Take ‘Page Likes’ or ‘Page Fans’ on Facebook (Meta) as an example. Page Likes help a brand as:

  1. Future ads can target Page fans directly or audiences similar to page fans and this increases conversion rates and decreases acquisition costs (hence improving the success rates of future campaigns).
  2. Page likes are an indicator to the marketer that people like your brand, its content, products or services. As such, they chose to like the page to support the brand and subsequently receive more content from the page. Marketers can use page likes as a strong indicator that they are moving in the right direction.
  3. A high Page Like count serves as social proof and tells people that you might be of some value at first glance.

But if marketers are measured based on the number of Page Likes they can get, it is almost too easy to work against the brand. I see brands purchasing Page Likes outrightly, whether from click-farms on Fiverr, offering cash and prizes to a broad audience, or creating paid tasks on Partiposts. The Page Likes from these sources would:

  1. Be unlikely to result in sales, because they may not be the right target audience if they are even human at all, and so resulting in a poor engagement rate. It will also confuse Meta’s algorithm as it considers the interests and demographics of your existing fans when delivering ads to new audiences.
  2. Be an unreliable proxy for how well your content strategy and product direction are working. These incentivised “fans” would dilute your real fan base.
  3. Lower people’s trust in your brand. A high follower count with few genuine posts or ad engagements indicates the brand probably bought followers.

Social media metrics as KPI targets encourage short-term thinking and bad decision-making. People will do whatever is necessary to not be seen as a failure, so you’re better off not using social media KPIs targets as a measure of success.

Beyond KPI targets

The focus on estimated social media results takes the attention away from what’s important – the business objectives. Here’s how you can plan without defining the KPI targets.

Translate business objectives into marketing objectives

Brand owners are never far from thoughts about sales, conversions, and downloads. But if a brand only runs ads for those ends, it will not be able to scale up; They would experience a marginally increasing cost per result for these Bottom of Funnel objectives because the accessible audience for your brand will run dry. A balance of Top Of Funnel objectives like brand awareness and engagement, Middle of Funnel objectives like leads, enquiries, traffic and sign-ups, is necessary for nurturing and growing the pool of accessible target audiences for a brand.

When we plan campaigns for our clients, we translate their business objectives into marketing objectives to lay out the scope of the range of marketing activities that would contribute to their business growth. We would also show the appropriate composition of content and ads to make their digital marketing efforts scaleable. This way, the campaigns proposed can be mapped towards a measurable outcome and marketers are aligned with the objectives of the brand.

Define the maximum cost-per-results allowed

Whether planning for these campaigns at the top, middle or bottom of the funnel, Marketers are rarely aware of the business’ profit margin. Instead, they work with a one-time or monthly “budget” so maximising results based on the budget and timeframe is their only imperative.

This is not an optimal approach. When it comes to customer acquisition cost (CAC), the lower the better. But identifying the highest acceptable cost is more important at the planning stage because it allows marketers to determine the composition of campaign objectives that would maximise your business objectives and the time it might take to get these results.

If you’re not in the know, here’s a trade secret: When you run ads on social media, the platform allows you to determine two out of three variables – metrics, budget, or deadline. For example,

  • Metrics-budget priority: You get to set a spending cap per result to get the right amount of metrics within your budget. The lower the spending cap the longer it would take to complete this campaign. For these non-urgent campaigns, we keep them running and track the estimated completion date based on the rate of spending/results.
  • Budget-deadline priority: You pay every time your advertisement is shown. By expanding the target audience, you determine the audience size such that you would finish your budget by a deadline. But by expanding your audience, you may dilute the pool of non-relevant audiences, and as such, the broader the audience, the lower the results. For these campaigns, we would pace the spending rate so that the audience is only broad enough to end the campaign on time.
  • Metrics-deadline priority: If you need results within a timeframe, you can adjust your budget and advertisement placements so that you get the results you need by your deadline. For these campaigns, the more ads ran and placements are allowed, the faster the metrics will be gotten. We would run multiple campaigns on more platforms to match the rate of metrics growth needed.

There is so much that a marketer can do to control your cost, speed of growth and results to ensure that your business stays profitable. I believe that every marketer will work to earn their keep, and ensuring that the brand stays profitable while bringing CAC down over time is naturally what they would do.

Improvement and Continuity

The perfect marketing plan probably exists, but the odds of it being fathomable is improbable. That reality should compel us to focus on starting first and improving thereafter.

To improve the performance of every campaign, we carefully define the hypothesis that we are testing for each campaign. By treating each campaign with scientific rigour, it forces us to consider the variables that would actualise better performance. Success and improvement, in this light, are only real if repeatable.

Secondly, we track not just the performance of each campaign but also customer behaviour. This helps us understand how they interact with our brand and their level of engagement. With this data, we can more accurately determine what messaging works at which stage of the funnel and for what type of audience.

Lastly, once a campaign is completed, we do a post-mortem to document what went well and what can be improved. This forms the basis of our learning that would feed into improving subsequent campaigns.

By constantly asking ourselves how we can improve performance, we get better at what we do – refining our skills, processes, and understanding of customer behaviour. In this way, we can hope to get closer to the perfect marketing plan. Or at the very least, get pretty darn close.

In summary

Using social media results as key performance indicators during the planning stages of a marketing campaign is futile for these three reasons. First, it’s difficult to accurately predict results. Second, even if you could predict results, customer behaviour is constantly changing and evolving. Third, post-mortem analysis is necessary to improve future campaigns. As marketers, we should focus on starting first and improving thereafter. By constantly asking ourselves how we can improve performance, we get closer to the perfect marketing plan.