Performance marketing, direct marketing or “direct response advertising” are frequently referred to as a “disciplined,” “measurable” methods of marketing. The reason for this is that, in a well planned and executed marketing effort, the cost of an individual promotion and the actual results of that promotion can be directly related to each other.

This precise information permits you to evaluate the productivity of your specific promotions in the past, and to apply that quantitative knowledge to the planning of your similar promotions in the future.

The result you get is – or should be – a carefully controlled allocation of promotion dollars to those efforts most likely to produce desired profit results.

Among the planning and evaluation tools you should use as a performance marketer are:

• The Profit and Loss statement (P&L statement; both a detailed format for precise profit measurement as well as a summary format for rapid calculations.)
• Contribution to promotion and profit analysis, including calculation of break-even response.
• Back-end performance analysis.
• Return on promotion investment analysis.

Each of these is discussed and demonstrated in the following sections.

The Profit and Loss Statement (P&L)

The Profit and Loss statement is a detailed record of sales and costs. It is important that you work one out in advance of a promotion (in which case it is called a proforma P&L) as well as after the promotion is over so that you can analyze the results.

Table 1 below shows what a detailed P&L statement might look like for a promotional campaign offer of an item selling for $29.95. For instance, this could be either a simple pay-per-click campaign at $0.25 a click, total 100,000 clicks or an SMS text campaign at $0.25 a text, total 100,000 texts. Either way, the conversion rate is 3.2% which gets 3,200 orders:

		Unit Value	Number of Units	Total Dollars
1	Selling price	$29.95	3,200	$95,840
2	Deferred payment price	–	–	–
3	+ Shipping & handling	$1.75	3,200	$5,600
4	Average gross order value	$31.70	3,200	$101,440
5	– Returns (10%)	$31.70	320	$10,144
6	Average net sale	$31.70	2,880	$91,296
7	Cost of goods per sale	$7.49	2,880	$21,571
8	* per unrefurbished return	$7.49	64	$479
9	Order receipt and processing
10	* order process and customer set up	$1.68	3,200	$5,376
11	* credit card processing fees (3.5%)	$1.11	1,600	$1,776
12	* installment billing	–	–	–
13	* customer service	$7.50	160	$1,200
14	Shipping and handling	$1.75	3,200	$5,600
15	Returns postage	$1.50	320	$480
16	Returns handling	$0.50	320	$160
17	Returns refurbishing	$0.75	256	$192
18	Bad debt, charge backs, bouncing checks (3%)	$31.70	86	$2,726
19	* collection effort	$1.00	86	$86
20	Premium (free gift)	$1.00	3,200	$3,200
21	Promotion (CPO)	$7.81	3,200	$25,000
22	Overhead	$3.80	2,880	$10,956
23	Total expenses			$78,802
24	Profit before taxes		2,880	$12,494
25	Profit % to Net Sales			13.7%

Explanatory notes for enumerated rows:

1. The price of the item is $29.95. In this example there are no deluxe options or accessories which might increase this price.
2. If there was a separate deferred payment price (higher than the upfront, normal price) it would be entered here. Then line 4 would show the weighted value of upfront and installment orders.
3. –
4. –
5. 10% of gross orders received and shipped are returned by the buyer.
6. Net sales are orders left after returns but before bad debt.
7. The cost of goods delivered to the shipping point is $7.49, Each sale incurs this cost. In addition, it is assumed that 20% of returns cannot be refurbished and therefore become an additional product cost related to each sale.
8. –
9. –
10. This line includes creating shipping label, customer database costs and manual verifictions if needed.
11. Assumes average credit card fee of 3.5%. Fee applies to 50% of orders that are placed as the other 50% are paid for by wire transfer and online payment services which charge the consumer for the transfer.
12. Since there is no deferred payment plan, there is no installment billing cost.
13. Complaints, inquiries, special requests by customers. Affects 5% of orders.
14. The actual cost of postage plus shipping department.
15. Seller is refunding return postage to buyer.
16. Cost of processing returned goods and paperwork.
17. Assumes 80% of returns can be refurbished and put back in inventory.
18. Some checks bounce and some credit orders will get charged back or receive a complaint. This amounts to 3% of net orders after returns.
19. Cost of following up on no-pays and charge backs.
20. The customer keeps the premium even if he or she returns the merchandise.
21. The promotion cost assumes 100,000 clicks or SMS texts at $0.25 a piece.
22. All other expenses (management, equipment, insurance, utilities) etc. not accounted for in items 7-21. The % comes from last year’s history or this year’s plan.
23. –
24. One additional cost that would affect this profit, but which was not allowed for above, is the cost of money – either interest on borrowed funds or the time value of internal capital.

The P&L in Table 1 above is very detailed and would be somewhat cumbersome as a means of quickly reviewing alternative marketing plans. Table 2 below shows a summary of some of the detailed expense items.

		Unit Value	Number of Units	Total Dollars
1	Selling price	$29.95	3,200	$95,840
2	+ Shipping & handling	$1.75	3,200	$5,600
3	Gross order value	$31.70	3,200	$101,440
4	– Returns (10%)	$31.70	320	$10,144
5	Net sales	$31.70	2,880	$91,296
6	Cost of goods	$7.49	2,944	$22,050
7	Order process, shipping, returns costs, customer service, credit card fees	$4.62	3,200	$14,784
8	Bad debt (3%)	$32.70	86	$2,812
9	Premium (free gift)	$1.00	3,200	$3,200
10	Promotion (CPO)	$7.81	3,200	$25,000
11	Overhead	$3.80	2,880	$10,956
12	Total expenses			$78,802
13	Profit	$4.34	2,880	$12,494
14	Profit % to net sales			13.7%

The numbers in Table 2 are all derived from Table 1. The simplification shown here consists merely of adding together the values for related expense items.

Another way of presenting P&L data you can use is in a per unit format, without reference to any actual total volume. This method can be helpful if you need quick evaluation of propositions prior to developing full-blown budgets for an actual campaign.

Note that the per unit values will be different depending on whether that unit is assumed to be an order (before returns) or a net sale (after returns). This difference is illustrated in Table 3. The values presented in that table are based on the information presented in Tables 1 and 2.

	Per Order	Per Sale (Assuming 10% Returns)
Selling price	$29.95
+ Shipping & handling	$1.75
Gross order value	$31.70
– Returns	$3.17
Net sales	$28.53	$31.70
Cost of goods (including unrefurbished returns)	$6.89	$7.66
Order process, etc.	$4.62	$5.13
Bad debt	$0.88	$0.98
Premium (free gift)	$1.00	$1.11
Promotion	$7.81	$8.68
Overhead	$3.42	$3.80
Total expenses	$24.62	$27.36
Profit	$3.91	$4.34
Profit % to net sales	13.7%	13.7%

Both the Per Order and the Per Sale figures are meaningful and useful depending on the analysis you want to perform. For example, you should evaluate the cost efficiency of order processing and promotion on a per-order basis. But when you want to analyze overall profitability, it is better that you look at expenses on a per-sale basis.

The P&L statement also allows you to examine the operating ratios of a proposition, i.e. individual items of expense as a percentage of net sales. An example of this is shown in Table 4 below.

	Dollars	Percentage
Gross order value	$31.70
Returns		10%
Net sale	$31.70	100%
Cost of goods	$7.66	24.1%
Order process, etc.	$5.13	16.2%
Bad debt	$0.98	3.1%
Premium (free gift)	$1.11	3.5%
Promotion	$8.68	27.4%
Overhead	$3.80	12.0%
Total expense	$27.36	86.3%
Profit	$4.34	13.7%

The ratios shown above would not be unusual in an offer of a moderate priced item in an average one-product landing page promotion. Cost of goods and cost of promotion in the mid-20% range and order processing, etc., in the mid-teen percent range.

For general merchandise, returns of 10% are typical (they can be much higher in apparel, late-night TV-type of products, or hard-sell promotions). The relatively low percent of bad debt in this example reflects the fact that half the orders are being charged to credit cards and there are no installments.

In offers of high-ticket merchandise, as well as in ecommerce, cost of goods is often in the mid-40% range, while promotion is in the mid-teens and order processing, etc. well under 10%.

Contribution to Promotion and Profit

The development of a proforma P&L allows you to determine the promotion response that will be needed to achieve break-even or different levels of profitability. This is done by calculating the contribution to promotion and profit that is available from each sale after all volume-related expenses have been allowed for.

In the example in Table 5 below, each $31.70 sale has total direct expenses of $18.68 – leaving $13.02 per sale for promotion and profit. This is equivalent to $11.72 per order assuming 10% returns.

	Total Dollars	Per Net Sale
Gross orders	$101,440
Returns (10%)	$10,144
Net sales	$91,296	$31.70
Cost of goods	$22,050	$7.66
Order process, etc.	$14,784	$5.13
Bad debt	$2,812	$0.98
Premium (free gift)	$3,200	$1.11
Overhead	$10,956	$3.80
Subtotal direct expenses	$53,802	$18.68
Contribution to Promotion and Profit per net sale	$37,494	$13.02
Contribution to Promotion and Profit per order (10% returns)		$11.72

The relationship between 1) contribution to promotion and profit, 2) profit objective and 3) promotion response is demonstrated in Table 6.

Campaign Objective	Profit per Net Sale	Available $ for Promotion (1) – Per Sale	Available $ for Promotion – Per Order (2)	Promotion Response Needed (3)
Break-even	0	$13.02	$11.72	2.1%
5% profit	$1.59	$11.43	$10.29	2.4%
10% profit	$3.17	$9.85	$8.87	2.8%
13.7% profit	$4.34	$8.68	$7.81	3.2%

Explanatory notes:

1. This is $13.02 – $ Profit.
2. Includes 10% returns.
3. This is $0.25 : Available $ for Promotion Per Order

If your campaign objective or minimum acceptable results were “break-even”, then all of the $13.02 available for promotion and profit could be spent on promotion. The result would be that all promotion dollars spent would be recaptured but there would be zero profit. To determine the break-even promotion response:

• Convert the contribution per sale into a contribution per order by subtracting the value of returns ($13.02 – $1.30 = $11.72)
• Divide the unit cost of the promotion by the available contribution per order ($0.25 : $11.72 = 2.1%).

Table 6 also shows the promotion response required to achieve 5%, 10% and 13.7% profit. The dollar figures shown in the “Per Order” column in Table 6 are the Allowable Costs Per Order at different profit levels, given the direct expense assumptions used in this example.

Back-End Performance Analysis

You have to analyze performance marketing promotions in terms of both front-end response and back-end performance.

Front-end response is the number of orders, the promotion cost (cost per click, SMS text, direct mail piece, etc.) and the resulting cost per order. In the example outlined above the response was 3.2%, the CPC was $0.25 and the cost per order was $7.81.

It can be dangerous to make future promotion decisions solely on the basis of CPO. For example, a lower CPO – say $6.81 instead of $7.81 – is not necessarily a more profitable opportunity.

The reason why CPO alone (the front-end) is not necessarily an accurate indicator of profitability is because back-end performance frequently varies as CPO varies. “Back-end” includes returns, bad debt, persistency (how long a customer stays active) and future value (response to other promotions in the future).

Increases in front-end response (and related decreases in CPO) are often promotionally induced – i.e. more intensive copy, offers, gifts, etc. are used to motivate the prospect to convert.

The additional business generated in this way may have lower back-end quality – i.e., it may have higher returns and bad debt or lower persistency and future value.

Therefore, only a complete analysis of front-end and back-end results can give you the answer to the question, “Which is the better promotion approach?”

The following example, Table 7, illustrates the impact that front-end and back-end results have on profitability. It also shows a method for determining what changes in back-end performance can be accepted for a given change in front-end response.

	Base Case	Response +50% & Returns Unchanged	Response +50, Profit at 13.7% & Returns ?	Sequence of Steps
Number of clicks	100,000	100,000	100,000
Cost per click	$0.25	$0.25	$0.25
Promotion cost	$25,000	$25,000	$25,000
Conversion rate	3.2%	4.8%	4.8%
Cost per order	$7.81	$5.21	$5.21
Order value	$31.70	$31.70	$31.70	1
Returns	10%	10%	27.4%	12
Net sale	$31.70	$31.70	$31.70	2
Per sale
* Cost of goods	$7.66	$7.66	$7.66	6
* Order process, etc.	$5.13	$5.13	$6.36	9
* Promotion	$8.68	$5.79	$7.18	10
* Premium (free gift)	$1.11	$1.11	$1.38	11
* Overhead 12%	$3.80	$3.80	$3.80	7
* Bad debt 3.1%	$0.98	$0.98	$0.98	8
Total cost	$27.36	$24.47	$27.36	5
Profit per sale	$4.34	$7.23	$4.34	4
% to Net $	13.7%	23%	13.7%	3

In the above example, the “Base Case” uses the same numbers as in Tables 1, 2 and 3. In the second column of figures, the assumption is made that response goes up 50%. Note that consequently promotion cost per sale decreases from $8.68 to $5.79. Since there is no change in back-end performance (an unlikely event) profit increases to $7.23 per sale or 23%.

The third column shows that returns could rise to 27.4% if response increased 50% and the marketer was willing to leave profit at the base rate of 13.7%. See the notes below to see how the calculation was done. Note that the same method could be used to analyze changes in other expense items as well.

Explanatory notes for last column:

1. Enter order value.
2. Enter net sale,
3. Enter % profit target. This can vary.
4. Enter resulting $ profit per sale.
5. Enter total expenses (net sale minus profit).
6. Enter cost of goods (the slight change in unit cost resulting from a different number of unrefurbished returns has been ignored in this example).
7. Enter overhead (12%).
8. Enter bad debt (3.1%).

9-11. These numbers are calculated by comparing the last two columns as follows:

Total expenses	$24.47	$27.36
Goods, overhead, bad debt	$-12.44	$-12.44
Other expenses (9-11)	$12.03	$14.92

$14.92 : $12.03 = $1.24 which is applied to values of items 9-11 in second column.

12. To determine permissible returns, calculate difference between promotion cost per order and cost per sale ($7.18 – $5.21 = $1.97). Divide that difference by cost per sale ($1.97 : $7.18 = 27.4%).

Return on Promotion Investment (ROP)

Another technique used by some companies to analyze alternative promotion opportunities is ROP – or Return on Promotion Investment, the return that is calculated in this instance is the contribution to overhead and profit.

The promotion investment that this return is related to is limited to media dollars that will be risked up front (premium costs are excluded because they represent a direct expense).

Table 8 below shows the calculation of ROPs for two situations. In the left column the response is 3.2%, returns are 10% and the premium costs $1. In the right column response is 3.8%, returns are 15% and the premium costs $1.50. In both cases, total up front promotion cost is $25,000.

Conversion Rate	3.2%	3.8%
Gross orders	$101,440	$120,460
Returns	$10,144 (at 10%)	$18,069 (at 15%)
Net sales	$91,296	$102,391
Cost of goods	$22,050	$26,185
Order process, etc.	$14,784	$17,556
Bad debt	$2,812	$3,355
Premium (free gift)	$3,200 (at $1)	$5,700 (at $1.50)
Promotion	$25,000	$25,000
Subtotal	$67,846	$77,796
Contribution to Overhead (12%) and Profit	$23,450	$24,595
Return on Promotion (Contribution : Promotion)	94%	98%
$ Profit	$12,494	$12,308
% Profit	13.7%	12%

The ROP calculation shows you that the right column represents a better return on promotion investment (98% vs 94%) even though the profit percent to sales is lower (12% vs 13.7%).

The advantage that the ROP approach can offer you is that it takes into account differences in contribution to overhead that result from applying a percent overhead factor to different sales volume assumptions. Often the overhead is not actually different, and the additional dollars should be treated as a profit contribution.

Cost of Testing

Testing new media and new traffic sources, new offers, new formats, new product categories, etc. is a vital part of performance marketing and direct response efforts. It can also be expensive. The reason for the high expense is that tests are usually conducted in relatively small quantities which results in high ad costs (i.e., generally the lower your traffic/ad/media volume, the higher their cost) and development costs.

In addition, many tests are unsuccessful, i.e. the response rate is less than would have been achieved with the control package. The combined effect of higher test promotion cost and lower test response rates is to lower profit on the over-all promotion effort in the short run.

Table 9 below illustrates this point.

Promo piece	Traffic Volume (clicks)	CPC	$ Cost	% Response	CPO	Profit
Control	100,000	$0.27	$27,000	3.5%	$7.71	+12%
Test A	10,000	$0.35	$3,500	2.8%	$12.50	-4%
Test B	10,000	$0.60	$6,000	4.2%	$14.29	-10%
	120,000		$36,500	3.5%	$8.69	+9%

In the long run, however, successful tests may lead to higher overall profits because more productive promotions have been discovered. In the above example, Test B has a response rate 20% higher than the control and, assuming the back end is as good, this will give the seller much higher profits when it becomes the control piece in the next campaign (at the current control’s cost per click, that is).

When analyzing test results, it is important to use the traffic/ad/media costs that will occur when the test is scaled in large, roll-out volumes. Table 10 shows the difference between actual and projected results with the Test B package.

Test B Promo	Traffic Volume (clicks)	CPC	Response	CPO
As Test	10,000	$0.60	4.2%	$14.29
As Control	100,000	$0.27	4.2%	$6.43

In this example, the projected CPO for Promo B is only $6.43 because, as the control, it would cost only $0.27 to promote per click. Of course, the actual cost is important in developing a campaign budget. But you have to use the projected cost when reading the results of tests.

Two-Step Campaigns

The examples used in the beginning of this guide all assumed a one-step performance marketing effort. Some companies use a two-step (or multi-step) method of selling.

The purpose of step one is to get a lead from the prospect. Step two is the effort to convert that lead to a sale. While the final P&L for a two-step effort looks much like that for one-steps promotions – there are two separate elements in the promotion expense that you should study independently.

The first is the cost of a lead and the second is the conversion rate of each conversion effort. Table 11 below outlines the economics of two-step marketing.

1	Lead ad cost CPC	$0.18
2	Response to ad	3%
3	Cost per lead (CPL)	$6
4	Cost per conversion effort	$0.40
5	4 conversion efforts cost	$1.60
6	Total cost lead + conversion	$7.60
7	Cost per order
	* with 10% conversion	$76
	* with 15% conversion	$50.67
	* with 18% conversion	$42

Explanatory notes for enumerated rows:

1. –
2. –
3. Cost per lead is ad unit cost ($0.18 cost per click) divided by response (3%).
4. Assumes that the conversion effort is made by an SMS test, a robocall, and two emails which cost on average $0.40 each.
5. It is customary to make several conversion efforts in two-step marketing situations (which makes it a multi-step effort in the end).
6. The total cost of acquiring ($6.00) and following up on ($1.60) a lead is $7.60.
7. The rate at which leads are converted determines the final cost per order. If 10% of leads costing $7.60 each are converted to sales, the resulting cost per order is $76. A follow-up conversion rate of 10% to 18% is not unusual with multiple efforts.

As the final CPOs in Table 11 suggest, two-step marketing is often more appropriate to high-ticket offers than to low-ticket offers.

In two-step marketing, it is important that you look at both the cost per lead and the conversion rate before making decisions about a particular campaign. A higher lead response (and therefore lower cost per lead) is not necessarily beneficial because it may result in a much lower conversion rate (and therefore a higher final cost per order).

Summary

1. Implement P&L, Contribution to Profit and back-end analysis in your day-to-day performance marketing decision making concerning your promotional campaigns.
2. Have appropriate live dashboards created so that you get all the necessary information in an automated manner.
3. Decreasing the front-end promotion costs may be counter-productive. Make sure you analyse your back-end, i.e. product returns, customer service and similar costs. Do this on a per-campaign basis, not lumping them together.