THE RKGBLOG

Promotional Economics

As we wade into what is hopefully the feeding frenzy of the ecommerce holiday season, we should take a moment to consider bidding strategies.

We’ve written extensively over the years about the importance of anticipating seasonal changes in user behavior, and of looking at that from the perspective of first touch, rather than last touch attribution. It’s worth noting, though, that holiday season often combines both changes in consumer behavior (higher conversion rates, often lower AOV, usually lower lifetime values of new customers acquired), AND promotional events that both impact user behavior and change the underlying economics of search.

Today, the focus is how we think about bidding during a promotional event in general.

Promotions affect paid search in a number of different ways:

  • Promotional Ad Copy:
    • Increases Click Through Rate, improving QS and bringing in some combination of more traffic and lower CPCs
    • Can pull in lower quality bargain hunting traffic that converts less well than average
  • Promotional Deals on the Site:
    • Increase conversion rates. Demand elasticity suggests that the same group of people will be more likely to buy when prices fall.
    • Impacts AOV. Can increase the shopping basket size or decrease it. Your mileage will vary based on vertical and type of promotion.
    • Lowers margin per sale dollar. Whether the AOV goes up or down, the marketing income per dollar of sale is reduced.

This combination of factors that can move the needle in either direction may mean that the smart marketer might either bid up or down, depending on how the anticipated values shake out.

Let’s throw out the algebra here for clarity:

BVPC = Base Value Per Click. The anticipated value per click not counting the promotion
CRD = Conversion Rate Delta. How much and in what direction will the conversion rate change because of the promo?
AOVD = Average Order Value Delta. If AOV is measured in sales dollars (rather than margin), then we also need the next factor, MPSDD
MPSDD = Margin Per Sales Dollar Delta.
PVPC = Promotional Value Per Click

It probably goes without saying, but just in case:

Combining influences is a multiplicative effect, not additive.

If we imagine a scenario in which the anticipated conversion rate bump is 50%, the expected AOV hit is negative 10% and the expected margin per sales dollar hit is 20% then…

This is the WRONG way to combine those effects:

This is the RIGHT way to combine those effects:

It’s also not hard to argue that a 10% off sale that cuts the MPSDD by 20% (assuming a 50% gross margin) isn’t likely to lift conversion rates by 50%. If we pencil in a 20% conversion rate bump we see that the PVPC drops 14% as a result of the promotion.

Does that mean we should bid 14% less in the case above?

That’s a harder question. If we want the advertising cost to margin ratio to stay fixed the answer might be yes. {Sidebar: While it’s true that we put dollars in the bank, not percentages, few advertisers I know can meaningfully move the margin ratio negatively and increase profits because they’re already pretty close to the bone in profitability.}

One might even argue that you should reduce bids more than that because the lifetime value of the incremental customers is probably lower than average.

But profitability isn’t always the goal. Creating buzz, driving the top line, ginning up brand awareness may be part of the objective of the promotion, and sometimes margin rates are deceiving. At the end of a season unloading inventory at a “loss” is often more profitable than the alternative. When inventory is perishable (hotels, tickets, food) insisting on profitability can be horrendously unprofitable. If the inventory is a sunk cost, any revenue is better than no revenue.

KEY TAKEAWAYS

  • Remember that promotions affect not only the AOV but also the margin per dollar of sale.
  • Consider whether the goal of the promotion is to drive more profits, drive top line sales, unload inventory, or some combination of the above.
  • Use historical performance data to anticipate these influences as best you can.

Have a terrific Q4 folks!

Comments
3 Responses to “Promotional Economics”
  1. Great education as always. Your blog is a constant inspiration not only to PPC consultant but also to educators such as myself.

    We are working on a PPC platform that involves simulations and dynamic scenarios. This would create an environment that would favour participative behaviour and would lead to higher level insight through active learning.

    Imagine the depth of discussions that could arise after using a realistic ad placement system such as Simbound.

    Your feedback.on our work will be most welcomed.

Trackbacks
Check out what others are saying...
  1. [...] Promotional Economics, Rimm Kaufman [...]

  2. [...] it has incremental cost. PPC, used inefficiently, tugs at your margins. (And for a quick look at at the economics of paid search and discounting, read George Michie’s blog post at the RKG Blog). So target carefully (tip 1&2), gather data (tip 5) so that you are driving a campaign that is [...]