[HP-35] Direct Mail Marketing of the HP-35 Pocket Calculator - WESCON 72

[Didier Lachieze]

The following article is part of the WESCON 72 conference proceedings. The 1972 Western Electronic Show and Convention was held from September 19 through 22 at the new Los Angeles Convention Center.

DIRECT MAIL MARKETING OF THE HP-35 POCKET CALCULATOR
Duke Castle
Hewlett-Packard Company

Imagine for a minute that it is Monday morning. The minute you arrive at your office, your secretary tells you that the president of your company wants to see you right away. You walk into his office and he is exuberant.

"Jack, I've just received the first prototype of our pocket calculator, and it's fantastic. It does everything we thought it would - logarithms, exponentials, trigonometric functions. It's a real electronic slide rule but it's much faster, more versatile and it's accurate to 10 places.

"And look", he says, as he slides the sleek unit into his shirt pocket, "It's a true shirt pocket calculator! We can sell it for under $400, just about a quarter of the price of comparable desk top machines."

"Jack, how should we market this product?"

What would you say? This problem was exactly the one that Bill Hewlett presented to his HP marketing people. To help you understand the thought process they went through in tackling this problem, let's briefly consider the development of the HP-35.

DEVELOPMENT HISTORY

After Hewlett-Packard established itself in the calculator business by introducing the Model 9100 desk top scientific calculator, several people began thinking about the development of a much smaller, portable version. The real impetus for such a machine didn't come, though until Bill Hewlett returned from a trip to Japan in the middle of 1970.

He had always been impressed with the ingenuity and innovation the Japanese electronic manufacturers had shown in penetrating new markets. He was quite impressed by the developments they were making in small, hand-held calculators, and he couldn't understand why American companies weren't leading this development since it was American semiconductor and packaging technology that made the whole concept possible.

When he returned to the United States, he asked HP Labs to explore the feasibility of a truly sophisticated portable calculator - something that did more than just add, subtract, multiply and divide. He was thinking of a machine capable of solving complex transcendental functions and yet small enough to fit in a shirt pocket and priced low enough that the individual engineer and scientist could afford it.

By February 1971 the study was complete and the report was that, yes, the machine could be made and the size and price goals met. With that, Bill Hewlett directed that the product be given top priority. And, he wanted it to be introduced to the public by the end of the year.

On September 17, 1971, prototype units were presented to the Board of Directors. In 11 months, HP had done what normally took 30. They were ecstatic. There was no question that this was one of the most innovative products Hewlett-Packard had ever developed. Now the question was - How do we market it?

MARKET CONSIDERATIONS

In approaching the question of how to market the HP-35, two things were considered - the position of the product in the market place and the most favorable channel of distribution. On the basis of function, it was apparent that the HP-35 fit nicely in the lower end of the sophisticated desk top calculator market. Hewlett-Packard was already quite active in the programmable desk top market. Its price, though, was low enough to place the machine within reach of individual engineers and scientists, thus making the buying decision a personal one. To profitably sell the HP-35 for under $400, units had to be produced in the thousands, with the majority of sales being initially to individuals rather than corporations.

In looking at channels of distribution, three approaches were considered - salesmen, retail outlets, and direct mail. The use of salesmen was the traditional way Hewlett-Packard sold most of its products. In fact, its success in the desk top calculator market was achieved through use of salesmen personally calling on customers and demonstrating the machine. This made a great deal of sense where the customer was relatively small and each sale was worth several thousand dollars. The potential market for the HP-35, though, was estimated to be 3,000,000 people. It was not possible for the present sales force to physically reach all of these people. Nor did it make economic sense to expand the sales force or create a new one to sell a $395 product.

The second approach considered was the use of retail outlets, or possibly third step distributors. Although this option was given careful thought, there were a couple of things against it. The first was the substantial wholesale discount that retail outlets historically received. This mark down ranges anywhere from 25% to 40% below the retail price and presented a marketing cost that far exceeded what it was felt could be achieved through other distribution methods.

The second consideration that eliminated this approach was the loss of contact Hewlett-Packard would have with the customer. One of the greatest assets HP possesses is its reputation for providing good customer service and support for all of its products. With many HP products, selling through third parties would push HP out of this valued close interaction of manufacturer and user.

The third approach considered was the use of direct mail. This approach quieted many of the objections cited above. It also promised some additional extras of its own. The first advantage was that direct mail enabled HP to deal directly with the customer. Any reactions to the product, favorable or unfavorable, could be detected right away. Also, any service or additional support required could be provided immediately. A second advantage was the ease of selecting prospects direct mail provides. Mailing lists are readily available for any market segment desired. In fact, they are broken down to such a level that it was possible to test very precise segments of the engineering and scientific community to determine what reactions our prospects might have to the product in general, and which specific advertising appeal they responded to most readily. A third advantage was the control of costs and orders. Since orders are directly related to the number of pieces mailed, if more orders are needed, more mail is sent out. If the backlog rose too quickly, then fewer items were mailed. The cost of using mail, of course, depends on the response rate. This number varies not only from product to product, but also from mailing to mailing. As will be shown later, though, the overall cost per unit appeared to be the lowest of all the approaches considered.

However, there is risk in anything, and some of the negative reactions to the use of direct mail concerned the fact that few companies had ever tried to market a complex, $395 product by mail. Who would shell out that much money for a product they had never seen? Could it ever be sold without a personal demonstration? What is the credit risk? Many companies had been badly burned by bad debt losses. These concerns all had to be answered.

WHAT HAPPENED?

After the September board meeting it was decided to introduce the product as soon as production capabilities could be established. In the meantime, a mailing was put together to test various engineering and scientific lists and a variety of types of promotional appeals.

The HP-35 was announced in January 1972 and the unexpected happened. The field sales force went wild. In a little more than three months the entire year's production had been sold and the majority of this had been through the Hewlett-Packard sales force.

Even though the production rate was being rapidly increased, delivery exceeded twelve weeks and sending out a mailing to solicit more orders did not seem to make much sense. So the test mailing was delayed until the middle of May. Everyone was anxiously awaiting the results.

DIRECT MAIL ANALYSIS

The most frequently asked question on the effectiveness of direct mail is: "What's the response rate?" Although it's a popular question, it is the wrong one. The real question is: "Was the mailing profitable?" Direct mail has been used to sell everything from $5 magazine subscriptions to $5,000 automobiles. Yet the cost of a mailing remains relatively stable. Thus the profitability of a mailing is determined by the revenue it generates instead of the response rate it receives.

Although it is not possible to go into the specific results achieved in the HP May mailing, it can be just as helpful to look at a hypothetical example.

Suppose we have two products we wish to sell by direct mail. Product A sells for $50 and Product B sells for $500. For the sake of argument, let us say that the manufacturing cost; i.e., direct labor & materials, is 50% of the retail selling price. Thus the direct cost of Product A is $25 and of Product B $250. It also means that each unit of Product A we sell we will contribute $25 to overhead and profit. The figure for Product B is $250.

TABLE I
MAILING BREAKEVEN ANALYSIS

Code:

                        Prod. A   Prod. B
                        -------   --------
1. Selling Price          $50       $500
2. Direct Cost             25        250
3. Overhead & Profit       25        250
4. Mailing Cost/ 1,000    250        250
5. Breakeven Sales        
   a.  Units               10          1
   b.  Revenue            500        500

If we assume a mailing cost of $250 per thousand pieces mailed, our mailing will breakeven if we can generate enough revenue to cover our mailing cost. If we use our overhead and profit contribution to cover our mailing expenses first, we see that we have to sell 10 units of Product A (see Table I) to break even, but only 1 unit of Product B to do the same thing. The interesting thing to note is that the break even sales level is the same for both products, $500. Thus we have a better feeling for the profitability of a mailing if we look at the revenue generated instead of the inquiry response rate achieved. Our analysis shows that this index is a better method of measuring the success of a mailing than the response rate.

Where response rate does have some value is in determining the marketing cost as percentage of the selling price. As one can see from Table II the response rate for Product A has to be much higher than Product B to give an equally acceptable selling cost.

TABLE II
SELLING COST AS A PERCENTAGE OF SALES PRICE

Code:

 Response     Cost/     Cost as % of Sales Price
 Rate         Order     Prod. A       Prod. B
 --------     -----     -------       -------
   .1%        $250       500%           50%
   .5%         $50       100%           10%
  1%           $25        50%            5%
  5%            $5        10%            1%
 10%            $2.50      5%             .5%
 20%            $1.25      2.5%           .25%

Assume total mailing cost = $250/1,000 pieces mailed.

The last thing that one has to consider when analyzing the profitability of direct mail is to consider some costs that are directly affected. One is the cost of processing an order. The standard order processing approach used within Hewlett-Packard is a very sophisticated computer controlled system that works very well for the rest of HP's product line but was too expensive for the HP-35. A separate order processing system was established that not only reduced the normal processing cost significantly, but also generated customer acknowledgments within 24 hours after receipt of the orders.

Another nagging cost is that for customer credit and bad debt losses. Individuals purchasing the HP-35 were given the option of sending a check or money order, or using their bank credit card. Companies purchasing were given Hewlett-Packard's standard terms and were subject to the usual credit checks. This approach has resulted in a very small bad debt loss.

A third cost area involves customer service and returns. The HP-35 is fully warranted for one year and all service is performed in the manufacturing facility in Cupertino, California. Field failures are mailed back to Cupertino, repaired and sent back to the customer within 48 hours. Considering the sophistication and complexity of the product, the failure rate has been extremely low.

As an inducement to try the product and get around the personal demonstration objection the mailing offered the customer the opportunity to try the unit for 15 days. If he was not completely satisfied he could return the unit for a full refund. It has been interesting to see that the unit is really self-explanatory and as result, the return rate has been infinitesimal.

CONCLUSION

In summary, the test mailing in May proved that the HP-35 can be profitably sold by direct mail. What still is yet to be proven is that these results can be duplicated in far larger mass mailings. However, given the very favorable results of the test mailing, we have every reason to believe that mass mailings will be equally successful.

In general, the original thesis that a $395 product can be sold through direct mail has been affirmatively answered. Our experience has substantiated the advantages of being able to target for specific markets and of selectively controlling order rates. It has also shown that direct mail is an exceptionally cost-effective channel of distribution.