Wrestling with an important decision or information management problem?
Need someone to help you think clearly through a tricky business issue?
Wrestling with an important decision or information management problem?
Need someone to help you think clearly through a tricky business issue?
In my previous article, we looked at how to estimate the profitability of an e-mail acquisition campaign, concentrating mainly on the mathematics required, and using average or typical rates where we needed them. Today we'll consider what steps we can take to increase the profitability of such a campaign.
Unless you are conducting a very small campaign, most of your costs are going to be for the rental of e-mail addresses, so I'm going to focus on those costs in this section.
Finding the company who will provide e-mail addresses to you most cheaply will certainly minimise your costs, but it's unlikely that it will maximise your profitability. Why? Because cheaper lists tend to be less targetted and so produce fewer purchases per 100 e-mails than would a more expensive list, and it's the revenue from purchases that you need to offset your costs and make a profit.
If you have a fixed amount of money to spend on a future e-mail campaign and you have identified a range of potential e-mail lists from which to choose, maximising the profitability of that campaign requires you to find the e-mail list that generates the greatest net profit (ie margin from purchases less costs) per dollar spent. The most reliable way to estimate which of a number of e-mail lists is the best, based on this measure, is to undertake a test campaign using each of the available lists. Imagine that you've done this for 5 different lists, with the results as tabulated below.
List# E-mails Delivered $ Purchases % Margin Cost per Delivered E-mail Net Profit per $ Spent 1 500 $1,200 25% $0.50 $0.20 2 1,000 $2,700 25% $0.60 $0.13 3 1,000 $4,000 25% $2.00 -$0.50 4 750 $2,000 25% $0.40 $0.67 5 1,000 $3,500 25% $0.50 $0.75
So, based on the net profit per dollar spent measure, List #5 is the profit-maximising list.
Note that List #3 produced a greater dollar value of purchases per e-mail than List #5 ($4 per e-mail for List #3 vs $3.50 per e-mail for List #5). This greater sales productivity is not sufficient, however, to make up for the relative expensiveness of List #3 ($2.00 per e-mail for List #3 vs $0.50 for List #5).
Conversely, although List #4 is the cheapest list at $0.40 per e-mail, it doesn't produce enough revenue to outperform List #5.
In reality, you might not have the time nor the budget to perform a series of test campaigns using different e-mail lists. In this case, there are some general things that you should consider in selecting an e-mail list to rent.
Even if someone who receives an e-mail from you does not immediately make a purchase, you should do everything that you can to form your own relationship with this potential customer. Have them visit and bookmark your website, or sign up for a catalogue or a newsletter. Make the most of the opportunity that you have to communicate with the people that you e-mail. Why pay more than once to attempt to acquire the same customer?
In the previous article, you might recall that we identified a number of steps that an individual reading an e-mail must complete in order to make a purchase. These steps were:
As we discussed earlier, the e-mail list that you select will have a bearing on the number of 'bounced' e-mails you encounter (and you should negotiate a maximum acceptable level), but there are some steps that you can take to minimise the number of e-mails that don't get delivered:
Next we need to encourage the e-mail recipient to open the e-mail. This is almost solely the job of the subject heading, so it's vital that you spend sufficient time - at least as much as you spend on the body of the e-mail itself - coming up with alternative headings until you find one or two that you think will work. If you're able, test a few different subject headings and identify those which elicit the highest rates of opening.
It's difficult to provide much general advice here, other than to point out that you should generally keep subject headings short (since people can vary the number of characters that are displayed in their e-mail preview window), factual, specific and compelling. For example, the heading "Save Money!" is less likely to tempt a spam-wearied recipient than say "Best-selling Books 80% Off".
Once the e-mail has been opened, we'll usually want the recipient to click on something to, say, visit our website. Bear in mind that there are still web users who are unable to click on any image links (ie links that are embedded in a picture or graphic) in the e-mail you send them because their e-mail client only reads text. These days, many people use text-only e-mail clients or set their e-mail clients to accept only text e-mail messages because it's possible for viruses to be sent via HTML e-mail. In any case, in my experience, text links are much more often clicked on than are image links, so you should stick to text links.
Also, recognise that many people download their e-mail and read it when they are not connected to the Internet, so links of any sort are of little value to them. You might consider including a return e-mail address or a phone number for these people.
Finally, of course, you want the recipient to purchase something. The e-mail should make it astoundingly clear what the recipient must do to initiate the purchase process - click on this button to visit the website, call this number, and so on. After that, the goal is to make the purchase process as simple and as seamless as possible - but that's a topic for another day.
posted at 7:36 PM
Today I want to look at the simple mathematics of e-marketing : what are the key performance metrics for an online campaign and how can you estimate campaign profitability before you launch? If you haven't defined success, how will you know what it looks like?
I'm going to consider an e-mail campaign designed to acquire customers, one in which you are renting 'opt-in' e-mail addresses of non-customers (that is, the e-mail addresses of non-customers who've given their explicit permission to be presented with marketing offers of the sort you're making. Run, screaming, from lists where the e-mail account owners haven't consciously opted-in; there's enough spam in the world already). To see examples of the sorts of lists that are available, see OPT IN INC, an excellent US example of a targetted e-mail list provider. For Australian examples, take a look at IDG Communications or EmailCash.
Firstly, let's consider costs. I'm going to assume that you are using your e-mail list provider to undertake the bulk e-mailing (often the per e-mail charge will include the cost of doing this). If, instead, you've built your own e-mailing capability then you should include a sensible portion of its development and running costs in the fixed and variable cost components that we discuss next.
So, what are the costs? Well, the fixed costs are for items such as copywriting and list setup. The defining characteristic of the fixed costs is that they remain fixed regardless of how many addresses you e-mail. The variable cost, or per e-mail cost, is whatever price you've negotiated for the rental of each e-mail address multiplied by the number of addresses you've rented.
I've seen it written that every equation in a book halves the book's potential audience. Still, I'm going to risk an equation here. Let's call the fixed costs F, the cost per e-mail e, and the number of e-mails you send N. Then, you've got:
Total Cost = F + e × N
Leaping back into the real world for a moment, you might ask : what's a reasonable price to pay for an e-mail address (that is, what's a reasonable figure for e)? The exact answer depends on the quality of the e-mail list you are renting - how targetted it is, how often people on the list respond to offers, and so on - but a price of $0.35-$0.75 per delivered e-mail is not uncommon for a clean, targetted list. More important than the price of each e-mail address, though, is its value to you. The best way to assess this value is to undertake a small campaign using a sample from the list and calculate what return you're making on each e-mail you send.
Now, what about the fun bit : how much can we expect to sell?
Think for a moment about the steps in the buying process with an e-mail. The recipient has to:
These steps are sequential : if the e-mail recipient fails to negotiate any one of them, then the e-mail generates zero revenue.
What's important is the percentage of the people that complete each step in the process. Let's give all these percentages a name:
Delivery Rate (D) is the percentage of the e-mails that you send that actually reach a valid e-mail inbox
Opening Rate (O) is the percentage of those people who successfully received the e-mail who then opened it
Click-through Rate (C) is the percentage of those people who opened the e-mail who clicked through to your site (or phoned the number you gave them in the e-mail)
Response Rate (R) is the percentage of people who clicked through to your site (or phoned) who subsequently made a purchase
So, the percentage of people who were sent an e-mail and subsequently made a purchase is given by :
% of Purchasers = D × O × C × R
Again, let me give you some real-world examples. Assume that :
Combining all these rates, we find that the percentage of people who were sent an e-mail and subsequently made a purchase is about 0.05% (1 × 0.5 × 0.1 × 0.01). That's 1 purchase for every 2,000 e-mails sent. So, if you're paying 50c per delivered e-mail, that's a $1,000 customer acquisition cost (ignoring fixed campaign costs). In some industries this is a reasonable figure, in others it's laughably high.
OK, let's bring it all together.
Say :
So your total costs were $11,000 ($1,000 + $0.50 × 20,000).
If your margin on each sale was greater than $1,100, you made an immediate net profit on the campaign.
Alternatively, if you can generate an average $1,100 in lifetime margin from the customers you acquired from the e-mail then the campaign has still been profitable. (The lifetime margin is the total margin over all purchases made during the lifetime of the customer relationship.)
Another way to look at profitability is as follows. Let's say a new customer is worth $500 to you. Imagine that you've just undertaken a test campaign using an e-mail list from company X and you've made 5 sales from an e-mailing of 5,000 addresses. If your fixed campaign costs are $1,000, what's the maximum you should be willing to pay per e-mail address for a future campaign using a larger sample of, say, 50,000 addresses from the same list?
Well, assuming you get the same performance from the bigger sample, your profit is going to be:
Value of each new customer × No. of new customers - Total Cost
which is:
$500 × 50 - ($1,000 + cost per e-mail address × 50,000)
You break even if this is equal to zero which, with a little algebra, you'll find is when the cost per e-mail address is 48c. This is the most you should pay per e-mail address.
posted at 11:51 PM
I'll be using this weblog to bring you fresh websites, news and information that's caught my attention.
For example, after I'd finished the website, I came across Infonomia, a website with great material on Information Management.
I've also stumbled across Alexa, a site that allows you to find the traffic-based ranking of any site on the World Wide Web. More interestingly, it also allows you to find websites that are similar to the one whose URL you input (it determines this by the fact that people who visit the site you input also tend to visit the sites that it displays as being similar). So, for example, if we input The Sydney Morning Herald, it lists (among others) The Australian Broadcasting Corporation, as well as The Times of India and The British Broadcasting Corporation.
This is a fantastic way to find websites that you might like : just input the URL of the sites that you often visit and see what similar sites it throws up.
I'm just going to try one more. Let's try Statistics.com. Alexa comes back with The WWW Virtual Library : Statistics in position #1. Looks to be a great site - see for example the section headed On-Line Educational Resources (you'll need to scroll down quite a bit).
Until next time.
posted at 1:24 AM
