Bob and Weave Click Fraud Thieves
By Neal A . Rodriguez.
I clicked the “Edit times and bids” link in the “Edit Campaign Settings” page in the Google Adwords interface to experiment with dayparting, as click fraud kept exhausting my client’s PPC budget while leaving conversion results unchanged. Some days I opened Outlook and see only one to three inquiries that counted as conversions on my Adwords account; I would open up Adwords and find my campaign cost at or above the $100 daily budget. Mike Moran and Bill Hunt write about dayparting in Chapter 14 of their book, “Search Engine Marketing Inc.” I never used it, and barely anyone writes content on it when you Google dayparting, so I figured it may be one of the new secret ingredients in a PPC guru’s sauce that they don’t want the competition to know about.
I searched my inbox for inquiries that arrived for the past month, and I noticed there were time periods when people submitted inquiries most on the site every week. For instance, I noticed that the first inquiry would land in my inbox a little after midnight on Saturdays, and more inquiries would fill my inbox for a period of about four to eight hours thereafter.
Thus I clicked the radio button next to my active campaign, and I clicked “Edit Campaign Settings.” I clicked “Edit times and bids,” and the Advanced Ad Scheduling page in my Google Adwords interface had a stack of green bars with checks in them. I clicked “Edit” next to the Saturday time bar and out came this drop down menu with each hour of the day listed. The first was set to “12:00 am (Midnight),” so I just changed the second one to “8:00 am,” when my inquiry influx would stop for about four hours. I configured the next time period by clicking “Add,” and out came the field with the drop down menu. I set the next time period to start at 2:00 pm and end at 5:00 pm, when I would get consecutive inquiries for the three hours. I set another time period up for the evening, and logged into Adwords later that evening.
I received about 3 more inquiries than I typicaly would on a Saturday. On Sunday, I switched the Ad Scheduling page to Advanced Mode, and I dayparted everyday, so my ad could be shown during time periods where I generally received most inquiries. I even changed the percentage of my bid that should be used when showing it. I finished adding time periods for each day of the week, and the stack of green bars on the Advanced Ad Scheduling page looked like a crossword puzzle when I was done; the dark green checks in the green boxes turned into dark gray X’s in light gray boxes, in between the shrunken green boxes that graphed the time periods when my ads were scheduled to show.
I thought the advanced configuration would use a percentage of my budget I specified, not my cost per click, so I set the time periods to percentages to add up to 100% of what I thought was my budget; e.g. Midnight – 7:00 AM 40%; 11:00am – 1:00pm 20%; 7:00pm – Midnight 40%. I set my daily budget to $100, instead of setting the budget to $40-$70 in the beginning of the day and then to $100 at the end of the day like I usually did; I would start the day with a partial amount of the budget because cyber-leeches would execute invalid clicks on select time periods too. Cyber-leeches seemed to click ads for a set time period, and stop when ads stopped showing. I would stop my ads when I stopped receiving conversions, setting the daily budget low, and resumed my ads, increasing the daily budget, during the next high-conversion time period.
I received 40 inquiries the first week I parted the day for a 42% increase of the 28 inquiries the week before. The next week I received 42 inquiries, and the week before Labor Day I received a 107% increase in inquiries – 58 – of the control week’s 28 inquiries. I have been leveraging dayparting since, adjusting the bid percentages, time period lengths, and daily budgets each day maintaining a weekly inquiry count within the high forties and low fifties.
At this time I don’t account for any sample distortion effects that can deviate my results by working with low conversion volumes; nor did I split-test to compress the time between the presentation of my control strategy to have Adwords show my ads intermittently throughout a day’s entirety and dayparting. I can say, nevertheless, that these inquiries have generated about $50K in revenue in a month’s time as opposed to less than $5K a month before. And for a startup every revenue hike counts.
[The above article is a submission for Marketing Pilgrimâ€™s Search Engine Marketing Scholarship Contest. Each Monday in October, entries will be published and the most popular article of the week will qualify for the $5,000 grand prize. If youâ€™d like to submit an entry, please view the contest entry-requirements and guidelines.]