The final day of SMX East 2016 has come to a close. We share our insights from the various sessions on Day 3. Even more SMX East insights from previous days below.
Now that SMX East 2016 has drawn to a close, we have our final day’s session insights to share. And now, the quick hits of Day 3:
Mobile PPC suggestions from “Your Mobile PPC Sucks (But It Doesn’t Have to)”
- One in four websites has not been designed to reach a mobile audience.
- Only 2% of companies say they have the capability to identify and measure mobile actions.
- Yet 37% of retail transactions are cross-device, while many companies still treat mobile and desktop as separate audiences, while ads containing phone numbers have been observed to drive 3x conversions.
- Voice search is also trending upwards - 42% of searchers surveyed have started using voice search in the last 6 months as of summer 2016.
- Local searches are also trending up; queries containing “near me” have increased 2x in the last year.
- Mobile continues to drive shopping transactions; 52% of shoppers surveyed plan to use a smartphone for holiday shopping before going to a store; 82% will consult their phone while in the store.
- It’s far to expect different targets and results for mobile versus desktop. For search marketers new to mobile, consider using the following Google-recommended directional mobile bid modifier to start.
[ (mobile conversion rate / desktop conversion rate) - 1 ] x 100% = Your mobile bid modifier
- Must haves for mobile ads:
- Be sure to use mobile sitelinks.
- Include key product details in your headline and D1.
- Landing pages should be fixed to be easier to use and faster to convert with as short a conversion path as possible - short forms with easy entry on pages that load quickly
- In short, “a drunk person should be able to convert on your mobile site.”
Retargeting suggestions from “Your Retargeting Sucks (But It Doesn’t Have to!)”
- Retargeting is essential to paid search, and has been an essential part of marketing for centuries - experts have suggested a minimum of anywhere from 7-20 touches before a product can “sell itself.”
- According to survey data from CommerceHub, the leading reason shoppers abandon their carts and ultimately do not buy products is that they decide “they don’t need it”
- Shopping cart abandonment has been observed to be most common for home goods, health & beauty products and electronics.
- Fortunately, most US-based digital shoppers (65% of males; 59% of females) do believe ads are properly targeted to them.
- Retargeting can, of course, take the form of traditional digital marketing, such as standard text ads (STAs) and display banners, or they can be dynamic, based on user behavior, browser history and abandoned shopping carts.
- Retargeting has been observed to have the highest rate of return about 24 hours after a shopper’s cart has been abandoned.
- Quick review of geo targeting hierarchy: Country, state/region, DMA/county, city, ZIP code, radius.
- Granular location segmentation can often provide performance lifts in search engine marketing (SEM). If running a nationwide US campaign, companies may see different and better results managing it as a 50-state campaign than a single-country campaign.
- A nationwide US campaign may miss as much as 2% of the total population due to the way Google segments states vs. countries. As such, it can be a good idea to consider unspecified campaigns to run in tandem with geo-targeted campaigns. In other words, a nationwide campaign optimized to include all 50 states, and an unspecified US campaign to include the entire US but excluding all 50 states, will likely pick up that potentially “missing” segment of the population.
- As of 2016, Google can track about 72% of impressions down to the ZIP code level, a 33% improvement since 2014. However, on mobile devices, only 8% of impressions may be tracked to the state level.
- Depending on a company’s business model, it can also be a good idea to consider optimizing for geo, based not only on interest level in the company’s product within a specified geo, but also to consider targeting any searchers located in the company’s chosen geo, as well as searches expressing an interest in the target geo itself. This practice has been observed to re-capture about 8% of impressions that would otherwise have fallen out of individual searches at the state level.
- Visual communication can improve conversion rates - according to an infographic shared at this session, on average, users just read 20% of paid search content. Visual information gets to the brain 60,000 times faster. Text paired with visuals increases comprehension further.
- Videos can also increase site conversion rates - they should be no longer than 90 seconds and feature a custom design.
- Infographics can increase time spent on a website. Long, scrolling infographics are less preferable than smaller “micro-narratives” - 95% of consumers prefer short-form content over long form. Again, custom images are more appealing than stock images.
- Text narratives should focus on the value provided by a company’s product and its benefits.
- Such narratives can start by focusing on the what (what the company’s produce does), then how (value proposition) and why (by creating an emotional connection with visitors).
- In all cases, emphasize trust. Do visitors trust your company’s promise? If the company’s value proposition is solid, but sales aren’t being made, trust issues may be the problem.
- There are 5 key components to image-based ads on Facebook:
- Calls to action (CTA) with strong messaging and colorful CTA buttons that “pop” out of the ad.
- High-resolution, professional-quality images for aspirational verticals such as retail and travel, or authentic-looking user-generated content for verticals such as service, whether authenticity and trustworthiness are important.
- Where applicable, video content between 7-23 seconds, with a maximum introduction sequence of 3 seconds; original animations work well for software and services; professional-quality stock footage for retail and travel; cartoon- or whiteboard-based animations for instructional “how-to” content.
- Use the “Rule of Thirds” - break out any image-based ad into a 3x3 grid of 9 blocks, and place your most interesting images at the intersections of these blocks to hold the eye.
- Showcase screen captures of product or application in action on its intended device/platform.
- Where possible, use custom photos in ads, not stock photos, particularly for landing page “hero images” - custom images have been observed to bring in an additional 5-23% conversions in the hotel vertical.
- If budget-constrained, consider seeking free web-based image editing and royalty-free, “free for commercial use” stock photos, or cheap-as-free tools and image websites.
Retargeting insights from “RLSA Will Save the Day”
- Pay-per-click (PPC) paid search ads face challenges today they didn’t face 15 years ago:
- Paid search CPC is at an all-time high in established markets.
- Desktop query volumes peaked three years ago.
- Conversion rates are low and don’t seem to be changing drastically.
- Standard paid search is arguably a matter of inserting your brand into a buyer’s journey at the right time, but doesn’t necessarily grow a company’s target market. It arguably just harvests existing demand from searchers who have already decided to buy.
- This is where remarketing lists for shopping ads (RLSA) comes into play - by targeting search ads only to people who have visited your site, companies in verticals such as shopping, tourism and education have observed 2-3x higher click-through rates (CTR). Since a higher CTR means a higher Google Quality Score, and Quality Score is inversely proportional to cost-per-click (CPC), these higher CTRS also mean a lower CPC.
- Brand affinity dramatically affects CVR as well - well-known brands tend to convert better.
- The biggest challenge to RLSA is its low search volume - RLSA ads target only users that know your brand.
- Google and Bing suggest using retargeting to adjust bids, customize ads and broaden keywords, but this might be overkill.
- There are new ways to use RLSA for very competitive markets with high CPCs, or for companies with smaller budgets:
- One tactic for low-budget companies is to set its entire search ad budget to 100% RLSA - this will arguably remove the waste of underperforming non-RLSA PPC ads.
- Another tactic is to again remove all non-RLSA PPC ads and allocate a portion of spend to RLSA, and the remainder to cheaper, awareness-based marketing campaigns. Companies that can get large amounts of leads into their top-of-funnel with PR stunts, viral videos or TV appearances can potentially do very well with this strategy.
- An additional strategy is “Super RLSA” - which incorporates Facebook Audience Insights to pull the demographic information of a company’s best customers, and mirror those patterns in bid strategy. This will ensure that all prospects are demographically aligned to your company’s product, that your company is remarketing to users that recently searched for your product, and by way of keyword targeting, your company is guaranteed to find these prospects while they have intent to purchase.
- RLSA has an additional unique characteristic. In markets in which companies can convince partners/competitors to do also run RLSA, everyone wins. Example: Rather than having multiple luxury car dealerships competing against each other for the same generic luxury car-related keywords and driving up the CPC for everyone, they can instead target recognizable, brand-specific keywords such as “Porsche,” “Ferrari” and so on, each paying only for highly specific, low-competition keywords with low CPCs and each only receiving highly qualified leads.
- This is why brand affinity is arguably the single most important factor that differentiates a good buying signal from a bad one.
Google Customer Match Musings from “The Lowdown on Customer Match One Year Later”
- Google Customer Match is the company’s recently released targeting feature that lets advertisers add the email addresses of customers directly into AdWords and target them as audiences.
- Customer Match lists must consist of at least 1,000 identifiable emails - outdated email addresses don’t count.
- Emails need to have been obtained firsthand via opt-in.
- Customer Match has no time limit in terms of the age of a customer touch - RLSA, in contrast, is now restricted to 540 days. However, it has been observed that customer match decreases in effectiveness with leads that are more than a year old, as customers may switch email addresses.
- Customer Match lists likely overlap with RLSA lists, and share the similarity that they, too, reach high-value searchers, but also share the limitation of reaching only a small sample size.
- Brands looking to incorporate Customer Match are advised to also check the demographics of their lists - older customers may use Yahoo! or AOL for email, while younger users almost all use Gmail. Brands should also be aware that .edu email addresses can cause problems with Google’s “identifiability” criterion, and are not recommended.
- Google is also working on a “Data Append” feature for Customer Match that appends postal address and phone number to email information - this is in very early stages but may develop into something much more interesting with more adoption.
- Modern advertisers are drowning in a sea of data that makes keeping sight of genuine goals difficult.
- A data sample of $250M spend, 80M website sessions and 12 months of data in retail and e-commerce showed Facebook to be the most improved player, making up 5% of all ad traffic (quadruple the previous year’s figure).
- However, the biggest amount of traffic predictably came from Google in a combination of organic search and paid search. Interestingly, email delivered 3x the revenue of all social channels combined.
- Mobile ad growth continues, but perhaps at the expense of desktop traffic; however, mobile has a less-developed attribution infrastructure, making it far less efficient of a platform for advertising.
- In any case, mobile traffic can be correlated to higher website revenue with higher-than-average CVR, likely a result of customers researching future purchases on mobile, then converting on desktop. This is also why cross-device tracking is now of critical importance to digital advertisers.
- Page load speed has also been observed to be a crucial metric for conversions - the average page load time is 6.5 seconds, and the average server response time is 0.76 seconds. Shaving off just 2/10 of server response time can lead to an observed 8% CVR lift.
- Bounce rate – Conversely, bounce rate has been observed to have little to no correlation to CVR. While a high bounce rate is traditionally viewed as a strong negative, it’s quite possible that in today’s age of short-attention-span shopping, a high bounce rate is instead indicative of customers quickly seeking qualifying info from your company’s website, such as its phone number, pricing or warranty details as they research a purchase.
- In summary, the two factors that showed the highest correlation to CVR are AdWords traffic and site load speed, while the factor with the lowest correlation to CVR was bounce rate.
About the Author
Andrew Park is a content marketing manager at QuanticMind. A UC Berkeley graduate and lifelong Bay Area resident, Andrew has done tours of duty in editorial, PR and marketing, and now works with the QuanticMind team to communicate the importance of data science and machine learning in digital advertising.More Content by Andrew Park