The problem with data is that a lot of it’s fake. Or inexact. Or just wrong. In the context of advertising, inaccurate metrics (faked, fraudulent, or otherwise) distort campaign performance, which might be why all those impressions you bought didn’t lead to any more sales conversions.

Lucidity helps fix that. In a recent campaign, our partner, a major entertainment studio promoting a global film release, needed greater transparency into their programmatic spend. Specifically, they needed help identifying which inventory sources were delivering actual value – meaning real conversions and not just more fake, inaccurate metrics.

Using blockchain verification, our partner was able to gain transparency into their campaign and increase their conversion rate by 214%. Here’s how it was done.

1. Identify Discrepancies in Reported Data

The key to separating what the numbers say happened versus what actually happened is spotting the discrepancies in what was reported. For example, let’s say the DSP you use reports having fulfilled 200.3 million impressions, but your SSP only reports 198.7 million.

What’s the truth? How do you optimize for what’s working and avoid what isn’t?

Usually, discrepancies are difficult to resolve and cost dozens of people hours to do so. They’re caused by activities that flourish in the absence of a transparent supply chain. Bot fraud is a well-known culprit, but other causes are just as prevalent, like behind-the-scenes auction games and inconsistent measuring and reporting practices. The bottom line is that, regardless of the cause, discrepancies are a signal for wasted spend.

Resolving discrepancies is a problem blockchain was built to solve. On-chain smart contracts programmed with a set of predetermined rules judge each campaign event (i.e impression or click) as either valid or invalid automatically, without the need for a biased, self-serving arbiter.

In the case of the film release campaign, Lucidity aggregated and cross-referenced signals from the programmatic supply chain (DSP, SSP, etc) to expose discrepancies in the incoming data. Exposing the discrepancies allowed us to verify whether or not every single impression and click actually happened based on whether or not the entire supply chain agreed that it did.

2. Optimize Away from Discrepant Inventory

If discrepancies are an indicator of waste, than moving spend away from sites showing high levels of discrepancy is a great way to eliminate poor placements and fraudulent inventory.

blockchain verification table

To optimize our partner’s programmatic campaign based on discrepancy rates, the following steps were taken:

  • The campaign was created and ran on the AppNexus DSP, using our partner’s standard site/app blocklist, pre-bid filters for viewability and brand safety, and AppNexus’s performance optimization algorithm.
  • Lucidity tracking ran alongside the campaign to aggregate event-level signals from the supply chain.
  • Lucidity eliminated any site or app that demonstrated impression discrepancy rates of over 10% and/or click discrepancy rates of over 20%

To test the impact of our optimizations, the campaign was split into two distinct phases. The first phase or “control phase” ran the campaign with only our partner’s standard list of filters and optimizations. The second phase, or “test phase”, ran the campaign with the standard filters and optimizations as well as with Lucidity actively optimizing for discrepancies using the thresholds listed above.

Important to note, the click discrepancy rate threshold is more forgiving by design due to the lower number of clicks and higher probability of anomalies.

3. Compare Real Numbers to Real Conversions

So how do we know it works? How do we know that verifying impressions and clicks using blockchain verification actually results in more real, valuable users?

To answer that question, our partner set up third party conversion tracking alongside Lucidity’s own impression and click tracking. This allowed us to compare blockchain-confirmed impressions and clicks with actual conversion numbers. In this case, conversions were confirmed visits to the film’s landing page.

The optimized test phase of the campaign delivered a 22% conversion rate, which was 214% higher than the un-optimized control phase conversion rate of 7%. This increase in conversion rate was achieved despite the test phase receiving approximately ⅓ fewer clicks than the un-optimized control phase. This suggests that the elimination of discrepant inventory also eliminated fraudulent traffic known to artificially inflate clicks.

Clicks and CTRs are widely used as a way to assess the overall performance of a campaign. But here we see that actual conversions went up as the CTR went down.

With blockchain verification, you have the transparency required to improve the conversion rate of your campaigns by ensuring that you’re only marketing to real, high-value users on high-value inventory. A few final takeaways:

CTR alone can be misleading: In an age where bot traffic is prevalent, clicks might not be the greatest indicator of success especially for in-app inventory. Clicks are easily imitated by computer programs, and therefore their impact on actual conversions or sales should always be scrutinized.

Quality over quantity: While the optimized portion of the campaign delivered about ⅓ fewer clicks than the un-optimized portion, it generated 214% more direct landing page visits. Having this data in hand would allow marketers to justify employing a quality over quantity media buying strategy, where buying fewer quality impressions leads to more conversions and more sales.

Transparency is a means to a greater end: While the word “transparency” has become a bit of a buzzword in the last year, it’s significant to see the actual tangible results that transparency can create. In this case, that result was a significant increase in conversion rate. But broadly, transparency promises the ability to know what works and what doesn’t.

Using blockchain verification, we can create a transparent view of a campaign’s marketing performance so everyone – from the advertiser to the DSP to the SPP and so on – agrees on what happened. When everyone agrees – campaign optimizations are easy, and advertisers finally have a complete understanding of their ROI.

That’s something we couldn’t do before in digital advertising. And that’s pretty cool.

To access the entire case study where Lucidity drove an overall 13% website traffic lift, please see here.