Anyone Do a Lot of Direct Mail and Want a Better Response Rate?

[Newby]

I'm not negative on it.

I think I've told this story before, but one time I was having a roof put on my house. I asked the roofer to give me the names of a few people he had put roofs on so I could see how happy they were. He gives me a printout with hundreds of names with addresses and phone numbers that he had roofed.

I figured I would work the list to market pre-need to them. I mailed a letter to the first 50 and called behind it to set appointments. I don't remember the exact stats, but it was a hot list. I sold a bunch of them.

All I knew about this group was that they owned homes, had enough money to pay someone to put a roof on their home rather than do it themselves, and all lived in my city or nearby. But they were a lot better than random names in the phone book. They were also better than random names in church directories which is what I usually used for marketing lists at that time.

You might find that pet owners are hotter prospects than non-pet owners. Or widows are better than married couples. Or childless is better than people with kids. Or people that like to order things from QVC, or almost anything could be slightly better than only screening age, income and zip codes. Who knows unless it's tested?

But if you don't have enough population, it won't matter anyway because you can't be so selective. Some agents already feel that women are hotter prospects than men. Could be true. But testing might tell you that widowed women are better prospects than divorced or married women are.

If you find a real niche that works it could be a real advantage for a telesales agent that has multiple states to target. I wouldn't dismiss it as nonsense at all. I'll bet Globe Life, AARP and colonial Penn test a lot of things we aren't aware of.

 

[senior-advisor-indiana]

I'm not negative on it.

I think I've told this story before, but one time I was having a roof put on my house. I asked the roofer to give me the names of a few people he had put roofs on so I could see how happy they were. He gives me a printout with hundreds of names with addresses and phone numbers that he had roofed.

I figured I would work the list to market pre-need to them. I mailed a letter to the first 50 and called behind it to set appointments. I don't remember the exact stats, but it was a hot list. I sold a bunch of them.

All I knew about this group was that they owned homes, had enough money to pay someone to put a roof on their home rather than do it themselves, and all lived in my city or nearby. But they were a lot better than random names in the phone book. They were also better than random names in church directories which is what I usually used for marketing lists at that time.

You might find that pet owners are hotter prospects than non-pet owners. Or widows are better than married couples. Or childless is better than people with kids. Or people that like to order things from QVC, or almost anything could be slightly better than only screening age, income and zip codes. Who knows unless it's tested?

But if you don't have enough population, it won't matter anyway because you can't be so selective. Some agents already feel that women are hotter prospects than men. Could be true. But testing might tell you that widowed women are better prospects than divorced or married women are.

If you find a real niche that works it could be a real advantage for a telesales agent that has multiple states to target. I wouldn't dismiss it as nonsense at all. I'll bet Globe Life, AARP and colonial Penn test a lot of things we aren't aware of.

Geico does it now. They only market to buyers. That's why they don't need agents. I'm not saying there isn't a way to get a more targeted list. I'm saying it's damn near impossible to figure out the buyers with insurance. That's why we cast a huge net, because we don't know what's currently happening today in Ms. Johnsons life that is going to make her respond.

We can mail the same 1k people over and over and the responses will be different every time.

 

[Josh]

Geico does it now. They only market to buyers. That's why they don't need agents. I'm not saying there isn't a way to get a more targeted list. I'm saying it's damn near impossible to figure out the buyers with insurance. That's why we cast a huge net, because we don't know what's currently happening today in Ms. Johnsons life that is going to make her respond.

We can mail the same 1k people over and over and the responses will be different every time.

Fair enough, but what you don't know is if all the responses are within the same group. So for example, if you're mailing the same list and you get different women 75, 78, and 80 responding, but it's always women 75, 78, and 80, it can get pretty clear what the trend is. I very much doubt it's that simple, but that's the idea.

----------

Well I very much highly doubt you have any experience doing this.

I don't have any experience doing this. I'm also not the one that's going to do this.

I'm shocked that you don't think others have done this or are doing this.

They don't have the data sets or the data science specialists. Most barely track what the response rate was on the age and income if they even bother to match it up with the records that'd be considered high-end, this is in another league.

 

[jdeasy]

If you sell lists vs selling insurance your opinions will vary greatly on what constitutes a good lead.

 

[Josh]

If you sell lists vs selling insurance your opinions will vary greatly on what constitutes a good lead.

The whole point is to revolve it around whoever is working the responses thinks is a good lead. Take a the entire list(s) you're mailing, get a solid breakdown on what each individual is, well beyond just age and income, then measure what the "good lead" is coming back (in this case an agent listing which one is good vs bad) and then use the metrics to see if the ones that are good share common elements. Maybe all of the responders are women 70-80 but the only buyers are 75-80 (to really oversimplify it). This would include a bunch of information. Hell, on homeowners we could even add the number of houses in the home. Maybe all of the buyers have two bedrooms vs 3+. Maybe male renters and female homeowners are the buyers. It's entirely possible that even if all this gets matched up and the numbers are ran the end result is a statistically insignificant difference between the groups in which case there is no marketable information here.

That said, most everyone on this thread is strictly thinking the senior market. This could easily be used against any number of other lines.

One thing that does seem very clear is that at least some people think it's a complete waste of time. Fair enough. I do appreciate the feedback, but what I'm really looking for is exactly what the thread was named, someone that currently does a lot of direct mail and wants a better response rate. If someone is doing their own mail and has their own information we could run it and see what happens. They could even run the mail business as usually, hand us the list they mailed and have us see if we can predict within a reasonable degree of accuracy which ones were the most likely to respond and which ones were the least likely. Even running this in reverse, if you carved out the 10%-50% that never respond to anything ever, that could help the ROI a great deal too.

In online marketing there are literally hundreds of ways folks do this. "You're wasting half your marketing money, do you know which half?" Most offline marketers are most are adding pURLs to pieces, I haven't seen anyone doing print mail really digging deeper into data. The thought process almost always revolves around the piece itself though and maybe some demographics on the data, I've never seen or heard anyone even talking about this approach.

 

[dstrack]

Geico does it now. They only market to buyers. That's why they don't need agents. I'm not saying there isn't a way to get a more targeted list. I'm saying it's damn near impossible to figure out the buyers with insurance. That's why we cast a huge net, because we don't know what's currently happening today in Ms. Johnsons life that is going to make her respond.

We can mail the same 1k people over and over and the responses will be different every time.

I can see your side of this statement. That said I think it is lazy/foolish to not test.

Rule number one any good marketer follows is "the control is the enemy". This applies to the target market and the DM piece.

Peoples emotions change due to environmental changes. This is always evolving. Your marketing needs to keep up with that. I understand for some this might be splitting hairs but thats how you get a higher response. Its free money if you do the work. What emotions triggered certain people to buy 5 years ago might be different now.

Newbys example was great. There are all kinds of similarities buyers can possess that have nothing to do with what would be considered "common sense" insurance triggers. Find these and you're king.

 

[dstrack]

Hey Josh how is this different then Epsilon? Can you tell me what categories of data they have to cross reference?

Thanks

 

[Josh]

Hey Josh how is this different then Epsilon? Can you tell me what categories of data they have to cross reference?

Thanks

I don't know much about Epsilon other than what I found on google. They're website is pretty vague about their approach.

I don't know all the details of what they're comparing against, but it's a large enough pool of different fields it should be very easy to spot trends if they exist. This would include traditional filters such as age, income, homeowner status, and a pill of others along with social media to the extent available. If it's something you or someone else is interested I can get them in touch with this other company and they can get into specifics. Just send an email to [email protected].

 

[dstrack]

Epsilon stores data from major manufacturer warranty cards, polls/surveys, magazine subscriptions, prior active buyers on DR offers, and the typical income, home, family size, cars, credit, etc...etc...

Not sure if they track social. That would be very cool. Especially if someone could take your list and cross reference the Facebook likes.

Do you know if they can do that? If so thats a gold mine right there.

 

[djs]

This approach will definitely work for 'impulse' sale type of things. Want to target a direct mail piece for a new restaurant? I'd bet you could come up with a decent list doing this type of analysis. In fact, many companies offer this type of analysis with their lists, or at least claim to.

With a 'life' purchase, I'm not sure you can track the data points fast enough to keep up with the trends. A few things jump out, such as when people turn 65, they may have an interest in medicare products. A lot of people will say when they turn 66, they have an interest again.

Thinking it through logically, for an impulse buy, like a restaurant, I don't really care if they get another coupon next week and try another restaurant, hoping they still come back to mine on occasion. So the fact they responded to my mailing and that was tracked and then resold to another company isn't a terrible thing.

For insurance products, once they respond to my advertisement, the model will say to sell that name over and over again so others can market to it and REPLACE the same product I just sold them. I wouldn't want that. In fact, a similar policy of internet lead vendors of re-soliciting them to get another quote every 6 months has not made agents happy.

Some verticals, this can work very well. Others, it will be a disaster. It will be interesting to see statisitics by vertical over time.

And Josh, this is old school stuff. It may be a new algorithm, a new way of looking at it, but tracking people by behavior has been done for a very long time. This is why google, yahoo, bing and everyone else wants you to install their toolbar on your browser, it lets them track what you do so they can market to you. Far more data than you ever get in direct mail responses.

Dan