You Can Glimpse Into the Future with Lead Scoring, But How Do You Make Money from It – HERE’S HOW

By Sean Fenlon on July 31, 2008


A Lead Buyer’s Perspective:

Please note that this blog post is limited to an overview of Lead Scoring from a Lead BUYER’S perspective. While there is significant overlap, many considerations and implications of Lead Scoring are quite different from a Lead SELLER’S perspective.

Part I – What is Lead Scoring

“Lead Scoring” is perhaps the biggest buzz in the lead generation and lead buying ecosystem. Lead Scoring was a significant topic of discussion at the TARGUSinfo Lead Quality Summit in 2007, LeadsCon 2008, and is expected to be one of the dominant themes of the upcoming 2008 TARGUSinfo Summit.

“Lead Scoring” refers to a process of assigning a score to a lead whereby the “score” is a prediction of the likelihood of a lead converting into a sale (or otherwise progressing to the next milestone of a sales process). At its core, the concept is similar to how a credit score predicts the likelihood of a borrower repaying a loan or how insurance companies now use scores to predict a consumer’s likelihood of filing a claim.

Lead Scores are often banded together into ranges in order to create scoring strata or categories (i.e. Red = 0-50, Yellow = 51-79, and Green = 80-100, or Class A through Class Z, etc.). This is often done to create equal amounts of leads in the bands when the scores are not necessarily evenly disbursed across a continuum of scores.

Lead Scores are determined by analyzing historical lead results relative to their data characteristics. This is fairly simple process when there is only one field of data to analyze. For example, let’s imagine for a moment that a lead was nothing more than an email address and our analysis indicated that 50% of email addresses with a Yahoo.com domain converted into sales while only 25% of email addresses with a Hotmail.com domain converted into sales. The lead “score” of all Yahoo email leads would be 50 and the lead “score” of Hotmail email leads would be 25.

Simple right?

Actually, it is, but only because we were only analyzing one field of data. The process becomes exponentially more complicated as we add more fields of data into the lead analysis. So, to continue our example, what if we added state and gender fields to our email field? If we look at gender in a vacuum, we’ll see that 60% of males convert into a sale and 40% of females convert into a sale. Now we look at the state field and find that the best state is Alaska with 70% of leads from Alaska converting into a sale and the worst state is Florida with 20% of leads from Florida converting into a sale, and all the other states scattered somewhere in between.

But how do we combine what we know about the email, the gender, and the state of the lead into a single composite Lead Score? A quick and easy answer is to say we should simply take the average from all three. Unfortunately, it’s not quite that simple as each field needs its own “weighting” in determining a single Lead Score. In other words, it’s possible that gender is a far more accurate field of data in predicting the likelihood of a sale than is the domain of an email address or vice-versa. Or perhaps the state the lead is from is the single most significant field in predicting a sale. “Weighting” the degree of influence of each field is the best method for normalizing these inevitable variations of influence.

Here’s a way to visualize this “weighting” concept. Imagine an Excel spreadsheet with thousands of records and the head of each column is the field name (Email, Gender, State, etc.). Now imagine a dial above each field name with values 1-100 and each dial had its own setting based upon how accurately that individual field was capable of predicting a sale and the sum of all the values on the dials was 100 (as in 100%).

Determining these dimensions of probability and the relative weighting of each individual field of data is not rocket science, but it is Calculus. The most common method for reconciling multiple fields of data in a single composite predictive score is multi-variate regression analysis. Many data analyst wizards can perform multi-variate regression analysis right in Microsoft Excel, but there are other more robust tools that also include this analysis as a standard mean-and-potatoes part of their offering including those of SAS, Cognos, and SPSS, which can make things a lot easier.

DoublePositive has been employing Lead Scoring using multi-variate regression analysis almost since inception. Our feedback loop, however, is not whether or not a lead will result in a sale (as we seldom have access to that data from our clients), but whether or not the lead will result in a LIVE Hot Transfer from our Call Center. For a lead to result in a transfer, it must pass through our DOUBLEconfirm™ process and only a small percentage of Internet leads will make it through this secondary layer of verification, qualification, and screening. Thus, Lead Scoring provides DoublePositive with a mechanism to prioritize which lead will be sent to our Call Centers in order to achieve the highest degree of Call Center Agent productivity and the highest possible lead transfer ratio.

From a data perspective, Internet leads are like snowflakes – no two are ever alike (notwithstanding duplicates). Thus, the more accurate the score for predicting outcomes the more valuable the process becomes. The key to the most accurate and useful lead scores are:

  1. Accurate and timely feedback loop data (back end)
  2. As much accurate lead data to analyze as possible (front end)

Without an accurate feedback loop of final lead outcomes, Lead Scoring cannot be performed. Most lead buyers will use a sale as the target outcome objective, but scores can also be optimized for other events such as an application being submitted or a LIVE Hot Transfer, which are milestones of a sales-process but occur before the actual sale. This is particularly useful with extraordinarily long sales cycles of 90 or 180 days or more. The quality of the score is highly-dependent on the accuracy of the feedback loop. A feedback that reports back only 50% all the leads that convert into sales will result in a Lead Score that significantly less than 50% accurate.

Accuracy of lead data on the front end is often limited to the accuracy of the consumer-provided data vis-à-vis an Internet form-fill process. However, companies such as TARGUSinfo and eBureau can use real time systems to validate the consumer provided data, and also append hundreds of additional fields of data to the lead (i.e. data about the individual consumer, but not provided by the consumer via the form fill). More accurate data and MORE FIELDS of accurate data on the front end can substantially add to the robustness and thus accuracy of a Lead Score.

Part II – A Lead Score is a Means to an End

Predictive Analysis though Lead Scoring via multi-variate regression analysis is both powerful and valuable. It literally provides a way to predict the future. But how do you MAKE MONEY from Lead Scoring.

Well, there’s the rub.

The predictive nature of Lead Scoring does not create any value unto itself – the Lead Scores must be utilized in some way in order to extract value from the process. There are two direc
tions in which value of the Lead Scoring process can be gleaned – upstream and downstream.

Upstream means addressing the deal with the lead providers from where the leads are purchased.

Downstream means creating new discrete business rules and processes that are triggered and dependent upon the Lead Score data.

Upstream would require a complete re-negotiation of lead buying terms and prices based upon lead score. For example, an organization could hypothetically demand from a lead provider a complete spectrum of lead prices that was directly correlated to the lead score. For example, I could offer to pay $1 per lead for the leads with a score of 1 and $100 per lead for the leads with a score of 100 instead of a blended average of 50 for all leads.

Of course, this re-negotiation is only beneficial if I can maintain the same level of quality and quantity while lowering the average cost. While there are certainly deals between lead buyers and lead providers similar in spirit to this example, most industry experts agree that the lead providers by and large are not ready or comfortable for this method of pricing. A blended average price per lead relative to quality and quantity still rules the day in the majority of lead buying deals. Moreover, a lead buyer’s scoring system does not directly lower the cost of the media necessary to generate the lead by the lead seller, thus any lower average price for the same lead volume will result in profit margin compression on the lead seller side – something most lead sellers will not be too eager to accept.

So, if re-negotiations upstream prove incapable of gleaning value from Lead Scoring via lower costs, looking downstream and creating new Lead Score-dependent businesses becomes the answer.

The most obvious process would be to put the leads with the highest probability of converting into sales into the hands of the most-capable sales professionals. This approach is not without its own challenges, however, as the leads from the bottom bands of Lead Scores will quickly kill the sales productivity of any sales professional as result of the time required to work through leads with such a low probability of converting into sales. In theory, the added productivity of high-score leads to high-performing sales professionals would be cancelled out by the lack of productivity by the lower-performing sales professionals being forced to work only low-score leads.

So, at first glance, it appears difficult to glean any degree of value from a Lead Scoring process either Upstream OR Downstream.

So the challenge is quickly identified – Lead Scoring is GREAT but how do I MAKE MONEY from it?

Part III – The Best Way to MAKE MONEY With Lead Scoring

The answer lies in first identifying the economic consequences of lead with low Lead Scores. Even in low Lead Score bands, there are “nuggets of gold” of leads that can be converted into sales. The problem is the amount of TIME required by a sales professional to mine the leads for the gold nuggets.

Time is the most valuable asset of a sales professional and sales organization and every effort should be made to allocate as much of their time into sales-related activity as possible. However, for a lead to be ready for sales process, LIVE contact must be established with the consumer, the consumer must be genuinely-interested in the product or service being sold, and the consumer must be qualified. Anything less than this is a waste of a sales professional’s TIME.

Leads with high scores typically require minimal time from a sales professional to establish LIVE contact (high-score leads are typically very responsive, thus the contact ratio tends to be substantially higher) and minimal time to assess the consumer’s interest and qualifications. Leads with low scores, however, can require excessive amounts of time from a sales professional to contact and qualify.

The key to making money with Lead Scoring is to find a process that minimizes the amount of TIME spent by a sales professional on low score leads.

A solution that is becoming enormously popular amongst lead buyers that have begun scoring their leads is to REDIRECT low-scoring leads to a LIVE Hot Transfer service provider to Contact, Qualify, and Transfer the consumer to a sales professional. Typically, the lead buyer is only charged for the leads that result in a LIVE Hot Transfer and the sales professional can utilize 100% of their time in sales-related activity with LIVE, genuinely-interested and qualified consumers as opposed to wasting their time chasing down consumers, leaving messages, scheduling call-back, sending emails, asking preliminary qualifying questions, etc.

A LIVE Hot Transfer service provider will utilize technology, and precisely-calibrated process, and low-cost labor to perform the time-intensive task of mining the nuggets of gold from within the leads with the lowest lead scores. Leads with the highest scores can be delivered directly to the sales professionals as usual, but as a result of limiting their lead flow to only the leads with the highest probability of converting into a sale, the sales professionals are able to make more sales as compared to when they are asked to directly work leads from across all Lead Score bands.

A typical outcome is a higher conversion rate of all leads into sales (sometimes 50%-100% higher), which can drastically lower the average cost of acquiring a new customer, even with the cost of the LIVE Hot Transfer service provider incorporated into the total average acquisition cost.

The diagram below illustrates this approach to creating different workflow processes based upon a Lead Score:

diagram

Part IV – Conclusion

Interest and adoption of Lead Scoring models is expected to continue to gain popularity. Companies such as TARGUSinfo and eBureau can provide complete soup-to-nuts solutions for Lead Scoring, including data validation, data appending, as well as the actual real-time scoring itself.

Hopefully, the readers of this overview can now see an immediate path to MAKE MONEY as a result of a process designed to take advantage of the predictive value of Lead Scoring.


Subscribe via Email

Archives

Top Hobbies & Interests

  • Music
  • Univ. of MD Men's Basketball (Terps)
  • Washington Redskins
  • Washington Wizards
  • Recreational (i.e. Don't-keep-score) Golf
  • Computers
  • Audiophile & Home Theater Equipment
  • Movies -- Just the Good Ones :-)
  • Fine Restaurants
  • Fine Wines
  • Sophisticated Stand-up Comedy

Recommended Books

Recommended Music

  • Pat Metheny Group -- The Way Up

    B0006M4SO6 B000I5XDWO B000I5XD10

  • Pat Metheny Group -- Speaking of Now

    B00005V6IF B0000E2W0U

  • Pat Metheny -- Orchestrion

    B008HFWF3U B008NR92ME

  • Pat Metheny Unity Group -- Kin (<-->)

    B00GZ6L48U 
  • Bobby McFerrin -- VOCAbuLarieS

    B00303WR3M
  • Don Ellis -- Tears of Joy

    B0006Q0XWS
  • Snarky Puppy -- aka "The Michael League Group"