Why Internet Leads Suck

BenC

New Member
3
I'm in marketing and looking for some advice/guidance from those of you on the sales side of the fence.

I have a real problem with the pay-per-lead model used by the leads industry. I learned to truly despise it while at eBay where it was responsible for such an extreme amount of affiliate fraud the FBI had to be called in. Its an incentive model that rewards quantity over quality and gives bad actors a big advantage over everyone else.

After catching a talk by a vendor who had the hubris to say there is nothing wrong with internet leads and that sales teams who can't work shared leads just shouldn't survive I thought I'd throw my 2 cents in.

You’re not imagining things, internet leads really do suck because the pay-per-lead model incentivizes fraud, plain and simple. Even without the fraud the lead model just isn't good for selling. For example, we recently had a chance to work with a major ad agency who handles the account of a large insurer ($100M+ ad budget). We develop predictive algorithms to optimize for different advertising outcomes. In this case we were able to offer the marketing team a choice:

A) Improve lead quality & increase overall revenue but generate fewer overall leads.

- OR -

B) Worsen lead quality & decrease overall revenue but generate more overall leads.

Guess which outcome the marketing team chose? (Answer: B)


After selling my SEM platform to eBay in '07 I had a chance to work with the team looking to tackle this exact problem for eBay. In 1998 eBay was focused on driving user growth in its marketing programs so it used the same objective when it started an affiliate channel in 2000. The program was designed primarily around generating new users. Managers of the affiliate program at eBay & cj.com were heavily incentivized to drive user growth. While they would take action on obvious affiliate fraud they never had an incentive to actually go and hunt for it. A hear no evil, see no evil attitude developed.

A number of bad actors realized that as long as they could maintain a minimum level of performance they could stuff their pipeline with low intent, incentivized or even fraudulent users. They grew to dominate the program because this model rewarded aggregators who can take high quality leads from A and mix in low value or even fraudulent leads from B,C,D to create a pipeline with overall quality X. They generated a larger volume of leads than everyone else and the name of the game became keeping pipeline quality just high enough to meet a minimally acceptable standards while generating as much volume as possible.

Its far more profitable to generate a low quality pipeline because low quality users/leads are easier to generate. The bad actors (aggregators) always had plausible deniability, "XYZ assured us they were generating traffic thru SEO, I'm shocked they were using a click farm in India to generate signups. We had no idea (wink/wink)". This model actually forced good actors to become bad just to compete - everyone had to focus on quantity over quality in a race to the bottom. You simply weren't rewarded for generating a high quality users.

The dollars involved weren't small either (think $10M+/yr affiliate checks) and it went on for years. All told, losses were probably over $100M (and that could be conservative).

Currently buyers have very little power in the marketplace and thats a problem, your only leverage is simply not participate and buy leads.

Ignoring internet leads really isn't the answer. Google has captured so much buyer intent that flows into SEO/SEM you just can't ignore this channel, especially if you want to grow. At the same time I don't see any way to buy leads in this marketplace without getting burned. Like eBay discovered, quality just gets worse over time until it all collapses in a race to the bottom.


There is a solution but it involves putting the power back into the hands of buyers who can then force some changes.

eBay discovered the solution to this mess is to move to a revenue based model. The price buyers pay for leads would be based on the revenue these leads generate. Since no-one in this industry is eBay this would need to be done in an open marketplace where buyers and sellers could transact and compete.

Under this model, buyers would compete against each other bidding a % commission of their sales. You "bid" based on the revenue you generate and this is translated by the marketplace into an effective price-per-lead for lead sellers.

This model correctly aligns everyone around common goal of increasing revenue/sales.

The main advantages of the performance model:

  • Buyers aren't antagonized by a lead refund process. Lead refunds are made redundant.
  • Sellers are getting actual market rates for their leads. Anyone who generates good quality leads should see 2-4X the revenue they see today.
  • Good actors are rewarded and have KPIs to dial in efforts and generate leads the market demands. Revenue focus, not lead volume.
  • No profit motive for bad behaviour, the bad actors either reform or disappear.
  • Any seller action that damages lead quality get punished by the free market. Selling a lead to call centres before putting it in the marketplace would simply result in a lower price per lead. Do anything to the lead that makes it harder for sales team to close and consequences are less $$. Waste the time of a sales rep playing games with bad leads and the market punishes you.
  • Marketplace provides buyers with predictable pipeline of ideal prospects and forecasting based on their bid. They get to see where they sit amongst peers in key KPIs to improve their sales process. This is where AI/Machine Learning can get involved.
  • Consumer experience is greatly improved.

To anyone who says the current model is working I dare them to fill out a few quote forms and see if they can keep their phone turned on. This isn't good for user experience (UX) and big G! (google) values UX above all else. If this industry doesn't reform we'll all wake up to a day where Google is doing leadgen. They've already been experimenting with this in other markets (plumbers etc). Geico already spends $100M+ on AdWords, guess who Google will be selling leads to.

I work with customer intelligence (CI) datasets all the time, there is no excuse for the lead return rates you are all seeing with the tools we have today to scrub bad data. In fact, every lead card you buy should come with a CI profile before you even get on a call. This probably isn't happening because sellers profit by trying to sneak bad leads by you, or worse sneak in aged leads that would be discovered using CI.

So, I'm curious. Marketing teams I work with don't see a problem or pain point to solve. They are all measured by lead volume so a different model is bad for them. I wonder if those of you who work in sales feel the same way. I've put together a survey below that has 1 question on it, if this marketplace existed today what would you be willing to bid for leads?

I have a bunch of Amazon eGift cards left over from another project so if your willing to hop on a 10 minute research call in exchange for $10 eGift card I'd appreciate learning about insurance sales. Maybe there is a problem to solve here, then again maybe not.

As new user I can't post links (so need to cut & paste):

goo.gl/forms/2MchsoylSbOAClrr2
 
What data do you think should be included in the CI profile?

The usual socio-economic indicators would probably be a good start. There was a really strong correlation between lifestyle/income and policy/revenue for the insurers we've worked with.

Past that, e-mail/phone can be matched to cookie pools these days, get an idea of where prospect is in the buying cycle based on their browsing history. I can think of a few additional consumer databases that could be matched against that could be useful.

Really, I think all of this could be used in purchasing leads (target a certain persona/customer value). In AdWords we've had alot of success optimizing for revenue just using zipcode bid adjustments, unfortunately with everyone seeming to be more interested in lead volume over lead quality I don't see much demand for it in marketing.

Curious, from sales perspective - what data would be useful in closing before getting on a call?
 
My company is trying to help make at least insurance lead buying process much better. My company is launching an electronic insurance wallet mobile application where the end user can upload all of their insurance policies onto their phones, or take a picture of their policies using the app, and in many states, we can get the consumer's insurance information directly from their carriers with consent. In Europe, this technology is becoming really popular with the insurance consumer.

We are still deciding how to best handle the leads on our platform. We are one of the only insurtech companies that are not looking to cut out the insurance agents and brokers.
 
While I would not argue with your basic premise, there are two issues that would have to be addressed.

1. To share revenue, commissions, the individual/entity receiving the revenue would have to be licensed in the state and line of business.
2. How to do account for carrier access and skill level? We aren't talking about a widget, but instead an agent reaching the lead and then selling a service. A service where what is received and the pricing can be all over the board. Not all policies are the same, no matter how much some people would like you to believe they are. Not only that, but a particular policy can have coverage levels and endorsements changed which can change the pricing.
 
BenC.
I don't post as much these days, but an agent friend asked me to take a look at your first post ever on Insurance Forums and give my thoughts/opinion. Since your post is a book, I don't have the time/energy to address every point at this time. You did a nice job, but I respectfully disagree with trying to compare retail leads on eBay to Insurance Leads. So I'll try to be brief.

I've been in the lead business since 2004 when I was an Allstate agent. Today I"m not active at HometownQuotes, but I do sit on the board.

In my opinion, one cannot correlate eBay marketing [tangible products] to insurance leads [intangible product].

If you sold insurance leads, would you base your success on the agent's ability to close a sale? Before you answer, know that over 90% of new Allstate agents don't survive 36 months. [at least that is what it used to be when I was an agent]. Insurance is not like a Retail business, product wise.

I've sold leads to all the major captive carriers [& a few non-captive] over the years & I know of not one that ever wanted Quantity over Quality. They all tracked ROI based on sales/leads. If we didn't meet their metric, they gave us a warning & we had 30-60 days to correct. If we could not, they fired us. Yes, I've had them ask for more Quantity, but never at the expense of Quality.
But hey, maybe we've not done business with the same carriers?

Call Centers are a different matter/breed.

VolAgent brings up good points.

I'm not going to write more at this time, but happy to respond to questions from anyone.

----------

Lastly,
I talk to agents every day. Happy to hop on a call with you & I'll donate your Amazon card to charity.

Contact number & email are in my signature.
 
BenC.
I don't post as much these days, but an agent friend asked me to take a look at your first post ever on Insurance Forums and give my thoughts/opinion. Since your post is a book, I don't have the time/energy to address every point at this time. You did a nice job, but I respectfully disagree with trying to compare retail leads on eBay to Insurance Leads. So I'll try to be brief.

I've been in the lead business since 2004 when I was an Allstate agent. Today I"m not active at HometownQuotes, but I do sit on the board.

In my opinion, one cannot correlate eBay marketing [tangible products] to insurance leads [intangible product].

If you sold insurance leads, would you base your success on the agent's ability to close a sale? Before you answer, know that over 90% of new Allstate agents don't survive 36 months. [at least that is what it used to be when I was an agent]. Insurance is not like a Retail business, product wise.

I've sold leads to all the major captive carriers [& a few non-captive] over the years & I know of not one that ever wanted Quantity over Quality. They all tracked ROI based on sales/leads. If we didn't meet their metric, they gave us a warning & we had 30-60 days to correct. If we could not, they fired us. Yes, I've had them ask for more Quantity, but never at the expense of Quality.
But hey, maybe we've not done business with the same carriers?

Call Centers are a different matter/breed.

VolAgent brings up good points.

I'm not going to write more at this time, but happy to respond to questions from anyone.

----------

Lastly,
I talk to agents every day. Happy to hop on a call with you & I'll donate your Amazon card to charity.

Contact number & email are in my signature.
I would have to agree with Bob. While I don't have the experience he does (been in the industry since 2012), I do know Bob personally and have worked for other companies where it echos the same sentiments.

Quality is ALWAYS number one, as that's what leads to a higher close rate for insurance agents. SaaS and insurance marketing run VERY differently than retail marketing, like comparing apples and oranges.

Agents now a days are looking for speed contact with interested parties, which is why you are seeing a push into calls, both warm transfers and click-to-call. Each of those have their pro's and con's.

Many companies these days will process credits to accounts when a lead is legit bad, but specific returns vary from company to company.
 
Thanks for the great points, VolAgent


1. Satisfying regulatory frameworks I think could be accounted for in how the purchase model is presented. It would be relatively easy to translate a "commission bid" into fixed marketing type cost. Its the end result I think everyone would be looking for, namely a performance based mechanism for lead pricing that accounts for backend revenue and alleviates the friction of buyers/sellers having to constantly negotiate lead refunds.

2. Sales being a skills based profession means any marketplace needs to have competition that takes this into account. In this model, agents effectively compete with their peers. Your overall "bid" would be a function of what % you are willing to offer of sales PLUS your predicted success score (driven by AI) for a given type of lead. Frank might be better at closing non-exclusive urban millennials looking for basic packages while a Lucy has greater success with exclusive high income suburban baby boomers which she upsells - they could both bid the same % but get very different leads in pipeline (and represent very different revenues for publishers). Lucy would always have an advantage over Frank for the leads she works best.

Agents new to the platform would have incentive of hooking in their CRM so system can learn their past success patterns and match leads to pipeline PLUS these new entrants could then be scored immediately to compete with peers. New agents or those without history would need to earn their way and be at a competitive disadvantage until they establish a pattern of success (in other words, they would be forced to bid more against peers until establishing themselves).

I'll expand on this concept along with carrier access in answer to Bob below.


While I would not argue with your basic premise, there are two issues that would have to be addressed.

1. To share revenue, commissions, the individual/entity receiving the revenue would have to be licensed in the state and line of business.
2. How to do account for carrier access and skill level? We aren't talking about a widget, but instead an agent reaching the lead and then selling a service. A service where what is received and the pricing can be all over the board. Not all policies are the same, no matter how much some people would like you to believe they are. Not only that, but a particular policy can have coverage levels and endorsements changed which can change the pricing.
 
There has been at least 2 bid model's in the past & neither are still in business. Maybe you have a better way to model it?

All the best to you in making it work! The insurance lead space can use the innovation for sure.
This space has consolidated so much since All Web Leads bought all the Bankrate properties at a fire sale about 18 months ago. [BR paid around $430M for 3 main brands in '09-'10 & sold for around $130m or so in less than 5 years!]
 
Interesting thread.

I know a few of the "bad actors" involved in the ebay scenario. There are always two and some times three sides to a story. eBay at the time was setting unrealistic growth goals and they knowingly allowed affiliates to use blackhat techniques to generate traffic.

I will digress from that topic since there are easily searchable court documents on that topic.

There needs to be a solution where a lead is generated and it can be tracked on who is dialing it etc. Another solution is rather than buy leads from an aggregator, require partners to drive leads to your website/lead form.

You need to be educated as possible with online marketing and lead generation if that is the gas that drives your engine.
 
Back
Top