Your Professional New Year's Resolutions for 2023

Here’s your professional New Year’s resolution:

Create and build a data driven lead generation pipe using continuous visitor validation to improve your digital marketing spend, and thus lower your CAC.

And now you might think: “Yeah, great idea! But, how do I know this will ever going to work?”

Let’s start with a real use case of a still happy customer. A few years ago one of our clients -at that moment being small startup- needed a fraud detection solution. Their #1 revenue source was their website and they did want to get rid of lead generation fraud and the increasing number of TCPA settlements. And thus they used our fraud detection to filter out all the bad stuff. And, guess what? Within 2 years of growth they were acquired for over $2 billion, which caused some commotion in the industry.

Such a success story is of course a unique case, but key to this is: Clean data! Without clean data your algorithms will be less effective, no matter how ingenious they are. The strength of clean data and clever algorithms enables you to continuously optimize and refocus your marketing campaigns by re-allocating budget to the best performers.

Hence this LinkedIn article describing how to setup and organize your digital growth in order to be more effective.

Marketing campaigns are executed with a goal. This can be creating brand awareness, increase MAU (monthly active users), attract visitors, generating leads, e-commerce sales. Bottom line is: in the long run marketing needs to increase sales and revenue and thus have a positive ROI.

If your website is your #1 revenue source, your path to conversion needs to be smooth and frictionless. Your metrics for the sales pipeline velocity (measured in $/day) have to be based on closed sales. This positions marketing as key-revenue driver, especially when future growth is expected to come from: even more marketing.

If you naively increase your marketing budget and thus your digital media spend, you will scale the problems that come with paid traffic accordingly, if not more. Scaling your digital media spend will result in attracting bots, click farms, lead generation farms, etc. burning your budget. This is mostly the result of cheap traffic being sold as premium.

“Half the money I spend on advertising is wasted; the trouble is I don’t know which half.” – John Wanamaker (1838 – 1922)

More than 100 years ago John Wanamaker didn’t know which half of his budget was wasted. His main reason, or better his excuse, was the inability to measure by which he created the ‘big excuse’. When you don’t reach your ROI targets, you just simply blame it on his infamous excuse. In fact, this statement has so permeated the industry, that it has become an accepted fact. But, hundred years further in time you are able to measure what is happening in your marketing funnel. The only question in this century is: Is the measured data real or fake?

In 2023 there should be no excuse for large discrepancies, only a few percent at maximum. This can be achieved by monitoring and measuring the traffic at your landing page. It enables you to measure the number of visitors having clicked on an advertisement, link in an email, etc. Some of these visitors will convert to a lead by leaving their contact details. But, as always, there’s a dark side. That’s why you have to measure for fake traffic and fraud.

This can be achieved with fraud detection like Oxford BioChronometrics’ SecureLead which provides real-time flagging of clicks and leads without false positives (FP) and false negatives (FN) giving you near perfect accuracy and thus making your campaign as efficacious as possible. In other words, with a fraud detection such as SecureLead you are able to cherry pick the genuine leads to be contacted and the genuine clicks to be retargeted. Anything flagged as fraud is either returned to its source free of charge, or being compensated as credit traffic.

Your landing page is where visitors arrive and are nudged to leaving their contact details or in e-commerce making a (digital) sale. This is the place where paid traffic arrives and needs to undergo a security check, before being processed. This is the ‘fraud checkpoint’.

This security check will be done in the background not affecting the user experience. It is achieved by dynamically adding a detection tag to the landing page, which will collect all sorts of information enabling you to know which campaign, source and the unique generated lead are flagged as fraud.

Knowing the fraud status at the visitor level enables you to filter out ‘bad leads’ and return these to their respective source and filter out ‘bad clicks’ and get credit traffic or rebate to compensate for this. At the aggregated level you’ll know which source provides good traffic and how each campaign performs, enabling you to re-allocate budget to better performing sources. It also enables you to know which visitors are genuine. This enables you to process the good clicks and re-target these visitors. The good leads are sent to the next checkpoint: ‘the DNC scrubber checkpoint’

At this checkpoint your leads are clean and free of “marketing-related” fraud. But, there is another nasty class of leads which can cause a lot of trouble. These leads have been filled out by humans, but are very risky to follow up. This type of leads is called the “serial TCPA[1] plaintiffs” and/or “serial litigator” leads.

When a visitor fills out your lead generation form and your call center contacts the prospect by calling or texting the provided number, they’ll claim: “I’ve never filled out your lead generation contact form”. And if you are unable to prove that they did, they’ll threaten to sue you. In most cases this ends with a settlement… but that makes it an expensive lead at a surcharge of minimal $500. At scale, this seriously affects your average customer acquisition costs (CAC).

To prevent calling known litigators and known serial plaintiffs you can check and scrub your leads against dnc.com’s ‘litigator scrub(r)’ API. This is the quickest and easiest way to mitigate your risk of a TCPA lawsuit, as roughly one-third of the TCPA lawsuits originate from a small group of repeat serial plaintiffs and litigators. If a lead passes this check it is ready to be conveyed to the next stage: the sales pipeline.

Your optimal path is based on genuine visitors converting to a lead and subsequently converting to a sale. Unfortunately, the majority of the visitors don’t convert and are thus eligible for re-targeting.

For traffic flagged as fraud it depends on your contract and the media type (CPC or CPL) you bought.
• CPC traffic should either be refunded as credit traffic or as a discount on future spend.
• CPL traffic should enable you to return fraud flagged leads at no cost.

Effectively this means you have solved John Wanamaker’s budget waste problem by cherry picking generated leads for follow up, and cherry picking clicks for re-targeting. Once you have contacted the leads you will know which ones did convert to a sale and which didn’t. This gives you an additional data set to be re-targeted at a later moment.

Campaigns and landing pages often run one or multiple concurrent A/B tests, where in the simplest case half of the audience gets option A (the changed version) and the other half option B (the standard version). For example, two different color schemes, different form layouts, etc. In order to know how successful the change (option A) was you merely compare the performance of conversion results belonging to option A to option B and historical conversion results.

Fraud by bots may be affecting by you A/B tests. For example, if a bot is programmed to fill out a form it will break whenever the form fields change. Especially when a single page form changes into a multi step form which is less intimidating to humans. The reason is that humans can benefit from these kind of changes, but bots however are simply applying a sequence of pre-programmed steps. Because bots are not smart enough to cope with unexpected changes it will lead to failure. If this happens and you wouldn’t be flagging and filtering fraud from your traffic, your A/B test will be skewed. Fraud in the control group will succeed, because the bots filling out the forms were programmed to this specific landing page. However, the test group may be more successful in the human group, but if the fraud% is larger than the difference between the test and the control group: you’re drawing conclusions on fake data.

Fraudulent leads generated by human operated lead generation farms on the other hand are not affected nor influenced by font sizes, color schemes or changes in form layouts. The only goal of this type of fraud is to generate as much (fake) leads as reasonably possible in order to claim the attribution. This type of fraud will only narrow the relative difference between the control and test group, because visitors are randomly assigned with a group and this type of fraud has no preference.

Your only solution to prevent this from happening is to filter out fraudulent leads upfront. This enables you to work with clean data and make decisions based on your real intended audience. The key to success is the quality of the fraud detection. Oxford BioChronometrics’ SecureLead detects fraud without false positives (FP) and false negatives (FN) giving you near perfect accuracy and thus making your campaign as efficacious as possible.

Once a lead enters the sales-pipeline you need to keep track of its status, ie. the conversion to a sale. In some cases this might take some time, but once you know a lead is converted, or not, you need to convey this answer to the data analytics. This enables you to calculate the customer acquisition costs (CAC). In addition, you’ll also need to monitor and convey similar data to the data analytics as soon as a customer leaves. This data enables you to calculate the customer lifetime value (CLV).

In order to calculate whether you are spending too much marketing budget acquiring new customers, or to conclude that your marketing campaigns are highly effective and/or your customer base is very loyal, you need to look at the CLV to CAC ratio. If this ratio is too low (ie. anything below 3:1) then your marketing effort might be inefficient and you’ll need to look why and how you can make your campaigns more efficient. Again, this is done through data analytics -excluding fraud- and optimizing your campaign, by looking at: what does work! This can be achieved by tying back the conversion metrics (both leads and sales) to the original campaign data enabling you to see what works and what doesn’t work. Based on that you might only re-target a specific group of your visitors instead of the whole group. It also enables you to optimize your contact forms, landing pages, etc. by running experiments and A/B tests and draw performance conclusions on real human data only.

When implemented correctly you will have proven that John Wanamaker’s ‘half of my budget is wasted’ problem can be solved and that this is something of the pre-computer era. You just have to measure and filter out the fake and fraudulent clicks and leads to be more effective in making the correct decisions. Oxford BioChronometrics can help you achieve this.