--- title: "Facebook Lookalike Audiences From Your Customer List" description: "Facebook lookalike audiences from your customer list: how to build one from a list you already own, clean it, set audience size and narrow to your trade area." canonical: "https://socialinfluencebuilder.com/en/blog/facebook-lookalike-audience-customer-list/" lastmod: "2016-05-22T14:57:00+00:00" lang: "en-CA" format: "markdown" --- HTML version: # Facebook Lookalike Audiences From Your Customer List > **Archive note.** This article describes the situation as it stood when it was published. The rules, tools and features it mentions may have changed since. Check the current information with the official source before acting. Facebook lookalike audiences from your customer list turn a file you already own into a way to find new buyers: you upload the list, Facebook studies who those people are, and it goes looking for others who resemble them. ## Facebook lookalike audiences from your customer list: what it actually builds Facebook lookalike audiences start from a Custom Audience you build by uploading a customer list. Facebook hashes that list, studies who matched, and finds new people who share their traits. The traits it reads are things like age ranges, interests, online behaviours and buying signals. The list does not need to be huge; even a few hundred matched contacts give Facebook enough to work with. The stronger and more consistent your customer data, the closer the lookalike tends to be. I think of it as teaching Facebook who already trusts my business. A lookalike is really a form of audience [segmentation](/en/glossary/#segmentation): you start from the people who already buy and ask for more of the same. When I run [Facebook and Instagram ads](/en/facebook-instagram-ads/) for a client, the lookalike sits on top of a clean list and a working pixel so the campaign learns from real customers. ## Customer list sources: loyalty sheets, booking files and reservation emails Most Greater Montréal businesses already hold a usable list without realizing it: a loyalty sign-up sheet at a Terrebonne boutique, a clinic's booking file, a restaurant's reservation emails. Before any of it goes to Facebook, confirm you collected it with proper consent. That matters because the people on the list are real customers who handed you their details for a reason, and the reason has to cover what you are about to do with the data. A booking form that only asked for a phone number to confirm an appointment did not ask permission to use that number for ad matching. ### Consent under PIPEDA and CASL before a list is uploaded Using customers' contact details to match them on Facebook is a use of personal information under PIPEDA, so the people on the list must have consented to that use. CASL governs the commercial emails you send them, not the upload itself. Before I upload any list for a client, I check that each contact knew their information could be used for marketing when it was collected. If I cannot point to a clear consent moment for a given contact, that person comes off the list before I build the audience. ## Custom Audience matching: hashing, formatting and bilingual duplicates Custom Audience matching depends on formatting and hashing the list before it leaves your computer. Hashing turns details like email addresses into scrambled strings, so Facebook never receives the raw data. Facebook's upload tool hashes the file in your own browser, or you can hash it yourself first. Formatting decides how many contacts match, so put one email per column, lowercase them, strip the spaces out of phone numbers, and add the country code with the 450 or 514 area code. For Québec lists, watch for bilingual duplicates: the same customer entered twice under a French and an English spelling of the name. Deduplicating before upload prevents wasted matches and gives Facebook a better chance of finding your customers. ## Audience size setting: a close match versus a broader reach The audience size setting decides how closely the lookalike resembles your source list. Facebook lets you move that setting from a tight match to a much broader reach. I usually start with the tightest match for my Montréal-area clients because it keeps the audience genuinely similar to the people already buying. Once a campaign has been running for a while and the ads are steady, you can test a broader setting to reach more people, accepting that the match loosens as the pool grows. ## Source list refresh as customers come and go Refresh the source list on a regular schedule, because customers move, emails change and new buyers arrive. I re-upload a client's list every month or two so Facebook keeps modelling from current customers rather than people who bought years ago. Export a fresh file from your CRM or email platform, clean it exactly the way you cleaned the first one, and re-upload it to the audience in Ads Manager. Facebook processes the new file and updates the lookalike. If the business is seasonal, time the refresh just before the busy period so the model reflects who is actually buying then. ## Location narrowing: from a Canada-wide lookalike to your trade area A lookalike is built for a whole country, so narrow it at the ad set level to the area you actually serve. The lookalike still describes people right across Canada; the location filter only decides where Facebook delivers the ads. For a brick-and-mortar shop in Laval, for example, I set a radius around the storefront so the budget reaches people who can realistically walk in. The lookalike keeps its quality; the location filter just controls where it shows. A business that ships across the country can leave the location wide, but most local shops should not. ## Lookalike-audience checklist for a Greater Montréal small business Run through a short checklist before you spend anything, because it catches the gaps that waste budget. - Export a clean customer list, one email per column, phone numbers with the country code and the 450 or 514 area code. - Confirm every contact consented to this use under PIPEDA, and drop anyone you are unsure about. - Deduplicate the French and English spellings of the same customer. - Upload the list, let Facebook hash it, and wait for the match to process. Facebook needs a source list of a reasonable size to build a lookalike. - Create the lookalike with the tightest match for Canada. - Narrow the ad set location to the suburbs inside your real trade area. - Set age and language filters to match your actual customers. - Check that your pixel is firing on your site so delivery keeps improving over time. Further reading: a similar project, [Sabbya](/en/sabbya/). ## Go further - **Service:** [Facebook and Instagram advertising that moves Montreal customers from scroll to sale](https://socialinfluencebuilder.com/en/facebook-instagram-ads/) - **Case study:** [Sabbya](https://socialinfluencebuilder.com/en/sabbya/) - **Glossary:** [Digital marketing glossary: Segmentation](https://socialinfluencebuilder.com/en/glossary/#segmentation)