How do you decide what to spend if your busi­ness is heav­ily dri­ven by events? For instance, you may be a con­cert ticket adver­tiser plan­ning for a major gig, or per­haps a sport­ing adver­tiser plan­ning for the upcom­ing 6 Nations Cham­pi­onship (a major rugby tour­na­ment in Europe). Do you sim­ply try to bid to as high a posi­tion as you can on your top key­words and hope that the impres­sions you gain will bring you suf­fi­cient con­ver­sions to make the spend worth­while? But is it really worth­while? How far will you be from your cost per order (CPO) goal? Will the aver­age order value (AOV) be the same, lower or higher?

The chal­lenge of events-driven campaigns

Events-driven adver­tis­ing cam­paigns are chal­leng­ing because:

(1) you know some­thing is going to hap­pen, but

(2) you are unsure the scale of the impact, therefore

(3) you are uncer­tain how to pre­pare for it, hence you imple­ment the “spend and wait and see” strategy.

As a result, you may find your CPO fluc­tu­ates wildly indi­cat­ing your bud­get had not been allo­cated in the most effi­cient man­ner (assum­ing con­stant AOV). The chart below illus­trates such a sit­u­a­tion. The red cir­cles high­light the instances where events took place and the adver­tiser bid up to such a level that CPO was much higher than what it was worth. The green cir­cles, on the other hand, high­light the occa­sions where no event took place and the adver­tiser pushed down the bids. As a result, the CPO was too low leav­ing demand to be exploited.


The myth of chas­ing impressions

The strat­egy many adver­tis­ers employ when man­ag­ing event-based cam­paigns may sim­ply be chas­ing the impres­sion vol­ume. When they can fore­see there will be a rise in search inter­est they plan to spend more, by rais­ing bids, to gain more clicks. How­ever, we will argue that the value of a click is the key fac­tor at play here instead. If the upcom­ing event only dri­ves impres­sions but the con­ver­sion rate and the aver­age order value (AOV) stay the same, then the rev­enue per click would remain unchanged and there is no value in rais­ing spend. It is only when an event dri­ves a change in con­ver­sion rate and/or AOV, where it is worth­while to adjust bids to spend more.

So if we know the value of a click is the key, then how do we actu­ally decide the bud­get on a given event-affected day? The answer is com­pli­ment­ing sim­u­la­tion with anal­ogy adjustments.

As Will Senior, an Account Man­ager at Effi­cient Fron­tier, com­mented “to ensure you always pay the cor­rect amount for each click, you need to accu­rately pre­dict how upcom­ing events will impact click value. The eas­i­est and most reli­able method to pre­dict a change in the click value is to use his­tor­i­cal data from iden­ti­cal or looka­like events. The click value change could come in many forms, from a con­ver­sion rate shift to cus­tomer value adjustment.”

Best prac­tice = sim­u­la­tion + anal­o­gous data

At Effi­cient Fron­tier we are proud of our sophis­ti­cated mod­el­ling and fore­cast­ing algo­rithms. We build click and rev­enue mod­els on indi­vid­ual key­words and bid on them at the most effi­cient lev­els given the client’s busi­ness goal. Our platform’s sim­u­la­tions allow the adver­tiser to estab­lish the spend level required to achieve the spe­cific CPO or ROI goal based on his­tor­i­cal data. When there are forth­com­ing events, whether a one-day sports event or the week lead­ing up to a major music fes­ti­val, we will also con­sider the pre­vi­ous sim­i­lar (anal­o­gous) events and their con­ver­sion rates/AOV in rela­tion to the aver­age con­ver­sion rate/AOV. The lat­ter ratio will pro­vide us a con­ver­sion rate adjust­ment fac­tor to build into the sim­u­la­tion that is auto­mat­i­cally gen­er­ated by our technology.

This could be illus­trated by the graph below. Let us con­sider only CPO for now and assume con­stant AOV. The blue line illus­trates what the sim­u­la­tion tells us for an aver­age day about the rela­tion­ship between spend and CPO of a cam­paign. When an event is to occur and sim­i­lar past events sug­gest that there is a higher con­ver­sion rate than nor­mal, the curve moves to the right (green line) so that we can increase the spend from just under $3,000 a day to over $4,500 and gain more con­ver­sions for the same CPO of $120.


This approach of com­bin­ing sim­u­la­tion with anal­o­gous data gives the adver­tiser con­trol in two ways for man­ag­ing event-based campaigns.

First, the bud­get can be deter­mined in advance and the adver­tiser sim­ply sets the bud­get for each day of the com­ing week or month in the bid man­age­ment plat­form rather than on the day.

Sec­ond, this reduces the daily vari­a­tion in CPO from the tar­get CPO. For our clients, on aver­age, once this struc­tured approach is imple­mented (from an intuition-based approach), the vari­a­tion of CPO from tar­get is reduced by 50%. The more sim­i­lar his­tor­i­cal events there are, the more you can refine your esti­mate of the adjust­ment fac­tor used to build into the simulation.

Even if your cur­rent bid man­age­ment plat­form does not read­ily pro­vide sim­u­la­tions, using past sim­i­lar events as ref­er­ence to your fore­casts for the upcom­ing event would still pre­pare you bet­ter than intu­ition alone does, as evi­denced by aca­d­e­mic research (e.g. Green and Arm­strong (2007)).

Apply­ing a more struc­tured approach, your “spend and wait and see” strat­egy can be trans­formed into “ana­lyze and antic­i­pate”. You’re in control.

Dr Wing Lee, Busi­ness Analyst