A crucial step when planning your campaign execution, improving your Facebook ad targeting and online marketing in general is to map which aim and objective Nosto’s Facebook ads should cover and which channels to use. This article introduces in a short summary what targeting options and channels you have alongside with audience-builder and basics how to use them.
Target Audience Glossary
To summarise, currently Nosto’s campaigns allow you to target following customer segments, which are listed here using Faceook’s own terminology.
- Core Audience as demographic targeting: Targeting new customers using following demographic settingsShows best selling products of a store and supports limiting these to selected ranges for each campaign separatelyAgeGenderLocationInterests [inside and outside of Facebook]
- Lookalike Audience as lookalike targeting: Lookalikes are people with similar characteristics to the customers you already have and people who might want to become customers
- To generate a lookalike audience, Facebook requires a seed audience of users. Nosto generates this automatically based on most recent buying customers.
- Lookalikes require at least 100 users who have interacted with your site from the region which you would like to target. Therefore Lookalike audiences and targeting is only available in those regions where you have had enough traffic and transactions.
- As audience’s size needs to be at least 100 users. In case number of transactions is low and there hasn’t been required minimal amount of conversions in past 30 days, Nosto supplements the figure with users, who have added something to their carts (e.g. buying intent)
- Shows best selling products of a store and supports limiting these to selected ranges for each campaign separately
- Custom audience as targeting site visitors: Site visitors re-targetingTargets customers who have visited and viewed one or more products during their visit. Nosto only targets users who are engaged and signaled their interest towards something specific, meaning you don’t target bounced or otherwise not-engaged audience.Shows products visitors have viewed during their visit and relevant alternatives to those, thanks to Nosto’s recommendation engineIncludes visitors who have visited a site in past 30 daysExcluding users who have bought something in the past 7 days
- Custom audience as targeting cart abandoners : Abandoned cart re-targetingTargets customers who have visited the site and added something to their shopping cartShows products visitors have added to their cart during their visit and relevant alternatives to those thanks to Nosto’s recommendation engineIncludes visitors who have abandoned their cart in past 30 daysExcluding users who have bought something in the past 7 days
- Custom audience as post-purchase targeting: Post-purchase re-engagementTargets customers who have recently bought somethingShows most relevant items related to previous purchase, thanks to Nosto’s recommendation engineIncludes customers who have bought something in past 30 days
Nosto continuously adds new and more advanced and flexible targeting options, which are announced on release notes pages.
Facebook automatically limits the frequency of ads shown to an individual user, whereas users who are already exposed to your ads, who haven’t engaged with them, are less likely to see them again as Facebook pays attention to relevancy of user’s newsfeed. A daily limit regardless of the campaign you create with Nosto is 2 per user and 4 if the user is fan / follower of your Facebook page.
Targeting Best Practices and Optimization
In this chapter we explore more optimization topics than general how-to.
Foremost think of your site and business as a whole. If you notice that more conversions occur on the desktop rather than mobile, whereas shopping starting point is commonly mobile, tailor campaigns, messages, budgets and audiences so that they support your business goal for example inspiring users differently on mobile: Instagram and Facebook, while possibly re-targeting them more aggressively on desktop.
Key target areas and markets make a big difference. If notable share of sales come from the UK, whereas Germany might be a new area for your business to conquer, adjust campaign strategy and budgeting accordingly.
Especially before launching prospecting campaigns, use Facebook’s audience insights to learn more about your key audiences. Don’t forget your own CRM data either as this might reveal interesting details about most valuable customer segments across different genders, age ranges and regions.
Last, consider is your target to drive new customers regardless of their interest (bigger reach and audience) or do you want to reach out only to a smaller segment of users (smaller audience, more targeted campaigns) interested of something specific your store has to offer. All these points will help you to refine what segments you should reach out to.
Customer acquisition (Prospecting)
Especially if you decide to target based core audience based on demographic settings, consider do you want to reach out to big share of users such as all women in a specific area or target a smaller user group interested of something specific, such as women between 20-40 interested in yoga.
As a rule of thumb and especially if you’re using lookalike targeting, don’t artificially limit the audience size, but only if there are clear business requirements or very solid reason pointed out by CRM data or audience insight data suggesting to remove a group of users. Facebook is very good at finding good prospects and lookalike audience is a great tool for the purpose. However, if you limit lookalikes also by interest, it’ll dramatically reduce the audience size, making the campaign reach smaller. Limiting lookalikes is recommended by interests when you can still create a meaningful size of an audience.
An audience of 1M users might sound surprisingly large and you might be poised to limit it, but it’s not advisable. As a rule of thumb, if Audience Insights or your CRM data point out a good limiting factor, you possibly should limit the audience. Also if you have well founded reason to believe that a certain age group shouldn’t belong to your audience, these could be excluded. The audience estimation is not the amount of users who will see your ad, but expected maximum share of users who might see your ad. In brief, Facebook chooses users who are more likely to click your ad and who are affordable according to your max bid settings, covered fully in an article here.
Campaigns with small audience sizes will suffer ad fatigue faster as there isn’t enough users to reach out to anymore, once they’ve seen your ad a few times. This is more common for prospecting campaigns as their audience size is often limited and defined when campaigns are launched and doesn’t in practice increase during the campaign lifetime. Thus, prospecting campaigns should by their nature run for a certain pre-defined period of time, before adjusting the audience, ad creative and other settings and then re-launching them again.
Don’t ignore a gender, even if for example a notable share of selection and shoppers would be women. You might want to run separate campaigns to each gender, by balancing budgeting based on expected revenues instead, and by refining products shown in the ads by gender. Base conclusions rather on testing than a gut feeling.
Lookalikes are a good way to navigate through different interests, if your goal is bigger reach, and you can’t define specific behaviorial interest based on your CRM or audience insights data.
As a performance optimisation tool, Nosto’s study of prospecting campaigns (new customer acquisition) shows that conversion-% doesn’t have a clear relation to return on ad spend metric (ROAS). This means that even though customers from ads might have way poorer conversion rates than your traffic on average, they might still deliver positive commercial impact compared to the campaign budget. Advertisers are often inclined to terminate campaigns, which have poor conversion rates, but especially with Facebook, pay more attention to ROAS as a key metric. Surely, if your conversion-rate is very low and ROAS also poor, it might indicate that your campaign targeting options and other parameters are wrong.
Nosto’s performance analytics are based on relatively short time period of 24 hours, meaning that new customers need to purchase the item nearly immediately after clicking or viewing an ad. A better insight to how your ads actually influence customer behaviour is to review different time windows under ad manager, where you can choose also longer time periods of 7 and 28 days or by tracking these in your web analytics using influenced conversions.
While purchase is clearly the target you should target as an advertiser, remember that it’s fundamentally a macro-conversion and a result of multiple micro-conversions. Instead of optimizing for macro-conversions immediately, an alternative and good online marketing tactic would be optimizing for email capture from prospecting campaigns and later to use email-marketing alongside with Facebook and possible other ad platforms to re-target customers who slipped away during their first visit to your site.
Re-targeting regardless of the used channel or ad platform is extremely important in modern e-commerce, where customers might browse multiple sites, using multiple devices before eventually making a buying decision. One of the reasons why Facebook’s re-targeting is very effective is due to its capability to target the same individual, even though they would use multiple devices, whereas regular cookie based services often read a user on mobile as a separate user if the user continues browsing on a desktop, and as a third user if they then switch to tablet.
Regardless where your visitors found out from your site and offering, a visit and especially viewing a certain product range or a product is a clear indication of interest, which Nosto and Facebook can use to deliver more relevant ads. Even stronger signal, intention to buy, is given when a user adds something to their shopping cart. These are the users which are very likely to buy while re-targeting ads help you to keep them engaged. Instead of targeting users who simply viewed your home page, Nosto’s visitor re-targeting is based on audience who actually viewed something within their interest, making it more effective than tools and campaigns targeting all visitors, albeit audience size is smaller.
According to Nosto’s analytics (Q4/2016), every cart abandoner who returns shopping after clicking or viewing an ad is worth of more than $14 (average visit value), having conversion rates as high as 16%. Due to their value, your bidding strategy should be adjusted accordingly by using higher bids towards the expected revenue a visitor might deliver. In short, bid for the true value of such a customer.
With re-targeting campaigns remember to consider first the audience size depending on your businesses size. As a small retailer, you might come across a chicken-egg issue, because if visitor volume is low, consequently the amount of users who abandon their carts is lower as only a small share of visitors eventually add items to their carts. A good estimation is to assume that Facebook is capable of identifying 60-70% of users who visited your site, but for instance if the number of abandoned carts is only 100 per day, it would mean an increase of 70 users a day for the audience. Following the same example, Nosto targets users who have abandoned their carts in past 30 days, meaning that you might be able to target only 2100 users a month max. Albeit valuable, the audience size might make these campaigns perform well in terms of ROAS, but the impact to business itself might be small.
Number of visitors who navigated to product pages is always bigger than those who left their carts behind, hence for a smaller site in terms of traffic volume, visitor re-targeting might deliver better and more meaningful results.
With re-targeting, the audience should be limited only if business reasons rule over the logic of ad campaigns. In addition, as products shown in ads are automatically personalized, there shouldn’t be a need to filter products unless some of them violate Facebook ad policy or you for any reason can’t advertise them on Facebook. Audience is already been engaged and shown interest towards your site and offering, hence neither of these should be limited, except of course regions where you don’t ship or if you run separate campaigns for each country.
As the old saying goes, selling to existing customers is cheaper and easier than to acquire new. Re-engaging existing customers is even more valuable than targeting cart abandoners, having average visitor value of over $17 respectively (based on Nosto Q4/2017 stats). Post-purchase re-engagement should be a tool in every e-retailer’s toolbox, but surprisingly many retailers are more inclined to target new customers, regardless of the proven impact.
Even if the logic might sound conflicting, best way to optimize post-purchase re-engagement ads is to increase the max bid as the value of an existing customer returning to shop is high. So even if you might pay a bit more for visitors as they’re more inclined to buy, you might still have extremely good return on ad spend for re-engagement campaigns with relatively low cost-per-acquisition.
Post-purchase re-engagement campaigns share the same logical deficiency with cart re-targeting as the size of an audience can only be those users who have bought something in past 30 days so sales volume might be low, but it should be noted that each sales increases loyalty and customer lifetime value.
Use of Instagram
Instagram is still relatively new as a marketing and ads platform, so the topic is given proper weight by exploring it separately. While there’s clear evidence that supports Instagram’s effectiveness as an ads platform (read more at Yotpo’s blog here), an advertiser for an e-commerce site needs to know a few key facts about Instagram and the use of Instagram as a platform for product ads before really seeing the boost on sales this channel offers.
Instagram is a mobile-only service and thus very effective in brining in mobile traffic. Mobile conversion metrics alone aren’t the best metrics to follow when evaluating performance as for example according to Google 60% of shopping journeys begin on mobile, but end in a purchase being completed using a full-screen size such as laptop. This suggests, that some advertisers might want to run prospecting campaigns more aggressively on mobile, while re-targeting could be focused on desktop and using other marketing channels like email.
Approximately 20 % of mobile app time is spent either on Facebook or Instagram, so leaving Instagram out of the equation is effectively a lost opportunity. Remember that Instagram is still a very different platform compared to Facebook – even though the same data is being utilized, same rules and targeting might not apply between Facebook and Instagram.
Use Audience Insights as a tool to validate your hypothesis and test if dual placement (sending the same ad to both Instagram and Facebook) works for your ads, or if you need to set up separate campaigns with different targeting options and marketing texts.
Last but not least, Instagram is a platform where users come for visual inspiration so adjust product catalogues used in your ads accordingly. Products should represent your brand story and the natural flow of inspirational images on Instagram.