Sign up for our daily recaps of the ever-changing search marketing landscape.
Boost Your Campaigns By Becoming An Expert Search “Forecaster”
In the Greater Seattle area, you never know what spring weather will bring — cloudy at 8 a.m., sunny at noon, pouring down rain at 5. I always start the day with a check of the forecast, so I know what to expect.
In the same vein, anyone planning a search advertising campaign should know what the day’s search “forecast” might be.
That’s because search isn’t a static thing. Search behavior changes each day, during the day as well as on different days of the week. Those changes can take the form of different user intent, different search volumes, different devices, or all three.
Advertisers can use search forecasting knowledge to sharpen their campaigns and increase performance. They can match ad supply with demand, change bids to meet changing search volumes, and take advantage of trends in mobile and tablet searches.
The Changing Search Landscape
Those of us deeply involved in search see every day how the landscape is changing. A few years ago, nearly all search was performed with a desktop or laptop PC. Today, according to Marin Software, 37 percent of all U.S. searches are performed on a smartphone.
What’s really interesting from a search advertising perspective is the behavior of mobile searchers. They aren’t searching — they’re hunting. Fifty-five percent of mobile search conversions occur within one hour of a mobile search, according to research conducted by Nielsen and Google in 2012.
These are searchers who can create instant business for the advertiser that connects with them and understands the evolving search marketplace.
A big part of that understanding comes from taking a look at how search takes place across the three primary devices most people use: PC, tablet, smartphone. Increasingly, the devices people carry with them — tablets and smartphones — see search behavior that varies from what takes place on a PC.
Daily Variations In Search
People’s activities change throughout the day and so does the context for their search behavior. For instance, we’ve seen that PC searches trend down during the afternoon and early evening (3 p.m.- 8 p.m.), as people get ready to leave work.
At the same time, tablet and mobile search trends up — people are using these devices on the subway, or out at dinner, or in the gym. You can see the trends in this chart, which are normalized for a user’s hour to factor in time zone differences:
Next, see below how volume shifts from PC to mobile devices and tablets during the evening hours and late nights. Increasingly, the late-night companion for people is not TV alone; it’s a tablet or smartphone.
Volume Trends By Vertical & Device
It’s also worth noting how search varies based on users’ intent, which also shifts during the day. For example, in the morning, most people have on their game face. They want to complete an important task, such as checking their brokerage or bank account. So we see financial services searches peak at 10 a.m. — right about coffee-break time.
Searches for travel — a much more entertaining pastime — peak in the evening, at around 8 p.m. You can see a sample of how those search trends vary by time-of-day, intent, and device across industries here:
Weekday Search Isn’t Weekend Search
Search shifts on weekends, as well. People are on a different rhythm — they’re up later, busy with families, away from their offices. Here, mobile and tablet devices have become a big part of consumers’ weekend lives, which ramp faster when compared to a normal weekday.
The PC on the other hand ramps slower on the weekend. You can see from this next chart how weekday and weekend search differ. Notice in particular how the search usage for mobile change:
Price Per Good Click
One last point about how search varies during the day or week is on return on investment — ROI. One of the things we measure is what constitutes a “good” click. In our view, a click that is not a “quick back,” meaning that the searcher spends more than 20 seconds on the clicked page, is a valuable click. I’ll call the cost of this good click PPGC — Price per Good Click.
As you can see below, which is based solely on data from Bing Ads, mobile tends to have a better PPGC throughout the day, reflecting the notion that mobile users are very serious searchers.
How Advertisers Can Profit From Search Patterns
The first thing to note, if you’re an advertiser, is that mobile cannot be overlooked — it’s simply too big a part of search. If you have not opted-in your campaigns to target mobile devices, you’ve opted-out of a terrific advertising opportunity.
Moreover, you need to understand how to price PC versus mobile ad bids. The PPGC tends to be lower on mobile devices, so in order for good clicks to translate into good CPAs (cost per action), an advertiser needs to have definitions of a CPA that are contextual for the mobile devices and mobile-friendly environments that support those actions.
As an example, many people today would not consider purchasing insurance through their mobile phone but would consider completing a contact form for more information. The context and therefore the primary metric for success should be appropriately accounted for when it comes to mobile traffic — not just in understanding the return on your search advertising spend, but how you develop your customer experience on your mobile landing pages.
Another takeaway is that understanding how user behavior changes with devices can help create better campaigns. As you’ve seen, behavior changes across devices, and between weekdays and weekends. To make the most of these changes, adjust budgets settings and consider differences between weekday and weekend to emphasize mobile when its use is peaking.
Finally, remember that mobile conversions could take the form of a phone call, a store visit, or a purchase. Include this in your ROI calculations in order to achieve the full value of mobile search advertising.
Opinions expressed in this article are those of the guest author and not necessarily Search Engine Land. Staff authors are listed here.