Posts tagged quant
I recently read an article about the money wasted purchasing contact names to build marketing databases (sadly, I can’t remember where), and it really hit home with me. At my current company, the previous Marketing VP was a huge proponent of buying massive, industry-specific marketing lists with the sole purpose of “building up our database.” Ah yes, the goal of all marketers…
However, as I read the article and fumed about the wasted money (between 50¢ and $2 per name for nearly 25,000 contacts), I had little more than gut feel to really know whether or not the lists did indeed provide marketing value. As a good quant marketer, I decided to throw together a spreadsheet to determine if the purchase of B2B retail and e-commerce lists was worth the price.
We purchased two types of lists: large, industry-specific databases from a list broker (between 700 and 9,000 contacts each), and smaller, campaign-specific lists from Jigsaw (between 150 and 1,000 contacts, targeted by job title, industry, and/or company name). For the larger lists, the list “universe” contained anywhere from 10,000 to over 1 million total contacts and we purchased a smaller, randomly-selected chunk.
Caveats: I looked at the purchased list data aggregated over the course of two years of email marketing, not on individual campaigns. Therefore, the individual campaign results for these may vary. The comparison numbers for targeted, non-purchased lists are for single campaigns.
Skimming 5% Off of the Top
First, I was curious to see how many of the 25,000 contacts contained invalid email addresses (not bounces, but simply bad emails) – essentially how much money was wasted out of the gate due to stale or incorrect information. It was just over 1,000 contacts, or around 4.3%.
After one list generated a 20% bounce rate (invalids, mailbox full, spam filter, etc.), our list broker gave two benchmarks against which to compare this (and I quote…):
- “The stated average delivery rate is 80%-95% so that’s within the norm.”
- “In today’s economy, a 20% bounce rate is not bad at all.”
Most list brokers will replace the invalid contacts with new ones, but looking at my results those replacements contained double the percentage of invalid contacts as the original lists, averaging close to 9% (albeit on a much smaller total). So not only is there a >4% shrinkage on the original list, it gets worse on the replacements.
Surprisingly, Jigsaw had a higher rate of invalid email addresses than the larger lists at 5.6%. Given the Jigsaw model, where users enter a contact’s details to earn points to “buy” other contacts, I’ve always naively expected the data to be cleaner and more complete. However, since Jigsaw doesn’t seem to validate the data, users are almost incented to enter bogus data.
But overall, our 4.3% average was relatively low, so I guess it’s to be expected and I can’t/won’t complain.
Since we primarily run email campaigns to generate leads, I looked at two metrics: any activity at all (web visit or email open) and sent-to-click-through rates. I used activity rate to give the lists the benefit of the doubt, assuming that the only way a contact would have visited our website was if they saw an unopened email and googled our name or just visited out of curiosity.
Of the eight large purchased lists, the best activity rate was 17% on a list of just under 1,000 contacts. The average, however, was only 5.5%. For the Jigsaw lists, the activity rate was a bit lower at 5.1%. Again, I’ve always been under the impression that the Jigsaw contacts were more robust, but this is strike two.
The click-through rate was surprising. Generally my email campaigns have a relatively good CTR, usually 10 – 50%, given the targeted nature of the content and the effective subject-to-content connection. If I can get someone to open an email, there’s a good chance that they will click through. Even better, my sent-to-click ratio runs in the 3 – 8% range.
The overall sent-to-click ratio for purchased lists was horrible, with just a 1.3% hit rate. Large lists were below 1%, with a few having zero clicks across a half-dozen campaigns. Jigsaw was much better, with a 2.6% sent-to-click ratio and approaching my results with non-purchased lists.
For some generic comparison, Jupiter Research reported in 2006 that untargeted broadcast emails (with no personalization or segmentation) generated an average CTR of 9.5%. Yes, this is CTR, not my sent-to-close ratio.
Worth the Money?
At this company, I’ve been able to drive an average lead gen cost-per-lead of around $20 across all marketing: email, events, and advertising. For email alone, it’s a bit higher at about $35 per lead. This includes purchased, rented, and organic contacts.
Looking at the price of the purchased lists over the total number of actual leads generated, the cost-per-lead is about $105, or 3x my overall average for email. Add to that the cost of increased spam scores and lost goodwill from unwanted emails, and even the cost of marketing automation (for which ours, Marketo, is priced partially on the size of your contact database).
Bottom line: This analysis convinces me that list purchases are a poor marketing and business decision, and I never should have listened to the previous Marketing VP. I guess this is just another reason why he’s no longer here…