Back to AcademyN°1 AI Starter Guide

AI & Marketing Glossary

The AI and marketing words you’ll hear every day, in plain language, each with a real example.

Marketing funnel: The steps a person goes through from first hearing about a business to buying and coming back: awareness, interest, decision, purchase and loyalty.; Example: Someone sees a bakery’s Instagram post (awareness), checks its menu (interest), reads reviews (decision), orders a cake (purchase) and orders again next month (loyalty).
Lead: A person who has shown interest and left a way to contact them, but hasn’t bought yet.; Example: Someone fills in the “Get a quote” form on a roofer’s website. They are now a lead.
Lead magnet: Something useful given away free in exchange for contact details.; Example: A gym offers a free 7-day meal plan to anyone who enters their email.
Call to action (CTA): The one thing you ask the reader to do next, usually as a button or a short line.; Example: “Book your haircut” on a barbershop’s website button.
Conversion: When someone does the thing you wanted them to do: buy, book, call or sign up.; Example: A visitor to a nail salon’s site books an appointment. That’s one conversion.
Conversion rate: The percentage of people who convert out of everyone who had the chance.; Example: 200 people visit the booking page and 10 book: a 5% conversion rate.
Landing page: A web page built for one offer and one action, often where an ad or QR code sends people.; Example: A QR code on a flyer opens a page about the café’s catering, with one “Request a quote” button.
Ideal customer profile (ICP): A clear description of the customer a business most wants: who they are, what they need and where to find them.; Example: A dog groomer’s ICP: busy owners of large breeds within 5 miles who want pickup and drop-off.
Value proposition: One or two sentences on why a customer should choose this business over the others.; Example: “Same-day phone screen repair in Worcester, or it’s free.”
Speed to lead: How quickly a business answers a new lead. Faster replies win far more customers.; Example: A contractor texts back within 5 minutes of a web inquiry and books the job before competitors even call.
CRM: Customer relationship management: one place to keep every lead and customer, their details, and every conversation with them.; Example: A salon’s CRM shows each client’s last visit, their stylist and a note that they prefer texts.
Brand voice: The consistent personality in how a business writes and speaks: its words, tone and what it would never say.; Example: A skate shop’s voice is casual and funny; a law firm’s is calm and clear.
Upsell: Offering a customer a bigger or better version of what they are already buying.; Example: A car wash offers the $25 wash-and-wax to someone who picked the $12 basic wash.
Customer lifetime value (LTV): The total a typical customer spends with the business over the whole time they stay a customer.; Example: A barbershop client who spends $35 every 3 weeks for 3 years is worth about $1,800.
SEO: Search engine optimization: improving a website and online presence so it shows up higher in free search results.; Example: A florist adds a page about wedding flowers in their town and starts appearing for “wedding florist Worcester”.
Local SEO: SEO for businesses that serve a specific area, focused on Google Maps, the local pack and “near me” searches.; Example: A pizzeria works on its Google profile, reviews and listings so it shows up when people nearby search “pizza”.
Google Business Profile: The free Google listing that shows a business on Maps and in search, with its hours, photos, reviews and contact buttons. It used to be called Google My Business.; Example: Search a café’s name and the box on the right with its hours, photos and “Directions” button is its Google Business Profile.
Local pack: The map and three business listings Google shows at the top of many local searches.; Example: Searching “barber near me” shows a map with three barbershops under it. Being one of the three brings in walk-ins.
NAP: Name, address, phone number. These should be exactly the same everywhere the business is listed online.; Example: If the website says “Suite 2” and Yelp says “Unit B”, the NAP doesn’t match. Pick one and use it everywhere.
Citation: Any place online that lists a business’s name, address and phone number, like Yelp, Apple Maps or a local directory.; Example: A plumber’s listings on Yelp, Bing Places, Angi and the chamber of commerce site are all citations.
Keyword: The words people type or say when they search.; Example: “Emergency plumber open now” and “cheap oil change near me” are keywords.
Backlink: A link from another website to yours. Links from trusted, relevant sites help search rankings.; Example: A local news article about a new bakery links to its website. That’s a valuable backlink.
Schema markup: A small piece of code on a website that labels details like the business name, address, hours and prices so search engines read them correctly.; Example: A restaurant’s site uses schema so Google can show its hours and price range right in the results.
SERP: Search engine results page: everything Google shows after a search, including ads, the map, AI answers and regular results.; Example: On the SERP for “tax preparer Boston”, the ads come first, then the map, then the websites.
Organic traffic: Visitors who arrive from free search results or unpaid posts, not from ads.; Example: A blog post about “how often to service your AC” brings 300 visitors a month from Google without paying for ads.
AI answers (AI Overviews): Summaries written by AI at the top of search results or in chat tools like ChatGPT, built from websites, profiles and reviews.; Example: Someone asks “best bakery in Worcester for birthday cakes” and gets an AI answer naming three bakeries with strong reviews.
Review gating: Asking only happy customers for reviews, or sending unhappy ones somewhere else first. Google’s rules don’t allow it.; Example: A text that says “Happy? Review us on Google. Not happy? Tell us privately” is review gating.
Review link: A short link from the Google Business Profile that opens the review box directly, so leaving a review takes one tap.; Example: A salon texts its review link after every appointment and puts it on an NFC card at the front desk.
Reputation management: Watching and improving what people say about a business online: getting reviews, replying to them and fixing the problems they reveal.; Example: A dentist replies to every review each week and changes the booking process after several reviews mention long waits.
Social proof: Evidence that other people already trust a business: reviews, ratings, testimonials, customer photos and client logos.; Example: A landscaper’s website shows “4.9 stars from 212 reviews” next to the quote button.
Pay-per-click (PPC): Advertising where the business pays only when someone clicks the ad. Google Ads works mostly this way.; Example: An HVAC company pays about $8 each time someone clicks its “AC repair” ad.
Cost per click (CPC): The average price paid for one click on an ad.; Example: $120 spent, 60 clicks: a $2 CPC.
CPM: Cost per thousand impressions: what it costs for an ad to be shown 1,000 times.; Example: A Facebook ad with a $9 CPM costs $9 for every 1,000 times it appears on someone’s screen.
Cost per lead (CPL): How much the business spends, on average, to get one new lead.; Example: $300 on ads brought 15 quote requests: a $20 CPL.
ROAS: Return on ad spend: revenue from ads divided by what the ads cost.; Example: $500 in ads brought $2,000 in sales: a ROAS of 4, or 4x.
Impressions: The number of times an ad or post was shown on a screen, counting repeat views by the same person.; Example: A post with 5,000 impressions may have been seen by 2,000 people, some more than once.
Reach: The number of different people who saw an ad or post.; Example: A post reached 2,000 people, and some of them saw it twice.
Targeting: Choosing who sees an ad: by location, age, interests, what they searched or whether they visited the website.; Example: A daycare shows its ads only to parents within 10 miles.
Retargeting: Showing ads to people who already visited the website or interacted with the business but didn’t buy.; Example: Someone looks at a boat rental page, leaves, and sees an ad for that rental on Instagram the next day.
A/B test: Comparing two versions that differ in one thing to see which works better.; Example: A café runs two ads that are identical except the photo, and keeps the one that gets more orders.
Engagement rate: The share of people who interacted with a post (liked, commented, shared or saved) out of everyone who saw it.; Example: 40 interactions from 1,000 people reached: a 4% engagement rate.
Short-form video: Vertical videos under about 90 seconds: Instagram Reels, TikTok and YouTube Shorts.; Example: A 20-second Reel of a barber doing a skin fade, start to finish.
User-generated content (UGC): Photos, videos and posts made by customers rather than the business. Always ask permission before reposting.; Example: A customer posts a photo of their birthday cake and tags the bakery, which reposts it with permission.
Content pillar: One of three to five main themes a business posts about again and again.; Example: A gym’s pillars: workouts, nutrition tips, member stories and behind the scenes.
Content calendar: A schedule of what will be posted, where and when.; Example: A spreadsheet with a row per day: platform, pillar, topic, caption, who makes it and when it’s approved.
Repurposing: Turning one piece of content into several for different platforms.; Example: One video about choosing a wedding cake becomes a Reel, a Google post, an email and three quote graphics.
Algorithm: The system a platform uses to decide which posts to show to whom, mostly based on what people watch, like and share.; Example: A Reel that people watch to the end gets shown to more people by Instagram’s algorithm.
Hook: The first line or first seconds of a post or video, which decide whether people keep watching.; Example: “Stop paying $80 for oil changes” as the first words of a mechanic’s video.
Opt-in: Clear permission from a person to receive marketing emails or texts. Texts in particular need written consent under US law.; Example: A checkbox at checkout: “Text me offers and reminders. Reply STOP to unsubscribe.”
Open rate: The share of recipients who opened an email. Some email apps make this number less exact than it looks, so clicks and replies matter more.; Example: 1,000 emails sent, 350 opened: a 35% open rate.
Drip sequence: A series of emails or texts sent automatically over days or weeks after someone signs up or buys.; Example: New gym members get a welcome email on day 1, a class guide on day 3 and a check-in on day 14.
Segmentation: Splitting a contact list into groups so each group gets messages that fit it.; Example: A pet store sends dog food offers only to dog owners, and cat offers only to cat owners.
Large language model (LLM): The kind of AI behind ChatGPT, Claude and Gemini: trained on huge amounts of text to predict the next word, which lets it write and answer questions.; Example: Asking Claude to turn your notes into a Google post uses an LLM.
Generative AI: AI that creates something new: text, images, audio or video.; Example: Writing ten caption ideas for a new menu item.
Predictive AI: AI that scores or forecasts, like which lead is likely to buy or who should see an ad.; Example: Google Ads deciding which searchers are most likely to call a plumber.
Prompt: The instructions you give an AI tool. Clearer prompts with context and examples get better results.; Example: “Write a 100-word Google post for Brancato Barbershop about our new Sunday hours, friendly tone, no emojis.”
Few-shot prompting: Including a few examples of what you want in the prompt, so the AI copies the style.; Example: Pasting three of the owner’s best review replies before asking AI to write a new one.
Prompt library: A shared, organized collection of prompts that work, so the team doesn’t start from scratch each time.; Example: A Google Doc with tested prompts for posts, review replies, emails and ad ideas.
Hallucination: When AI states something false with confidence, like a made-up statistic, quote or source.; Example: AI writes that a bakery “has won Best of Boston three years running”. It never has.
Context window: How much text an AI can consider at once, including your instructions, anything you paste in and its reply.; Example: You can paste a whole month of reviews into most tools today, but a 300-page manual may be too long.
Token: The small pieces of text AI tools read and write, roughly three-quarters of a word each. Usage limits and prices are counted in tokens.; Example: A 750-word blog post is roughly 1,000 tokens.
AI chatbot: A chat window on a website or app that answers questions automatically, and should hand off to a person when needed.; Example: A rental company’s chatbot answers “Do you deliver to Logan Airport?” at 11 p.m. and collects the lead’s phone number.
Human in the loop: A person checks and approves AI work before it reaches customers.; Example: AI drafts review replies each morning; the owner reads and approves them before they’re posted.
AI assistant: An app or website you talk to in plain language, such as ChatGPT, Claude or Gemini. You ask for something and it writes, explains, plans or creates it.; Example: A plumber asks an AI assistant to turn three bullet points into a polite quote email.
AI model: The trained "brain" inside an AI tool. One company can offer several models: faster ones for quick tasks and more capable ones for harder work.; Example: A salon owner switches to a more capable model in their AI assistant to analyze a year of booking data.
Training data: The text, images or other examples an AI learned from. It shapes what the AI knows, and why it may not know recent events or your business.; Example: An AI assistant without web search doesn’t know about the bakery that opened last month, because it wasn’t in its training data.
Multimodal AI: AI that works with more than text: it can look at photos, listen to audio, read documents and sometimes create images or speech.; Example: A landscaper snaps a photo of a yellowing lawn and asks the AI assistant what might be wrong.
AI image generator: A tool that creates a new picture from a written description, or edits a photo you upload.; Example: A café asks an image generator for a watercolor of a rainy morning to use as the background of its new menu.
Deepfake: AI-made video, audio or images showing a real person doing or saying something they never did. Making one without permission can be harmful and illegal.; Example: A scammer clones a business owner’s voice to call their bookkeeper and ask for an urgent transfer.
Voice AI: AI that listens and speaks in a natural voice. It can answer phone calls, book appointments, take messages or read a script aloud.; Example: A dental office’s voice AI answers after-hours calls, books cleanings and texts the front desk a summary of each call.
Knowledge base: The written facts an AI assistant is allowed to answer from: hours, prices, policies and common questions. Better answers start with a clear knowledge base.; Example: A moving company adds “We don’t move pianos” to its knowledge base, and its chatbot stops promising piano moves.
Transcription: Turning speech into written text. AI does it in seconds, so meetings, calls and voice memos can be searched, summarized and turned into to-do lists.; Example: A contractor records a site walk-through on their phone and gets a written punch list from the transcript.
AI agent: AI that doesn’t just answer but takes steps for you, such as searching, filling in forms, booking or working across apps, usually with your approval at key points.; Example: An AI agent finds three available caterers for a Saturday event, compares their quotes and drafts an email to the best one for the owner to send.
Workflow: A set of steps that run in order, often automatically, to finish a task.; Example: New web lead → add to CRM → text the owner → send the lead a welcome text.
Trigger: The event that starts an automation.; Example: A new form submission, a new booking or a new 1-star review.
Action: A step an automation does after the trigger.; Example: Add a row to a Google Sheet, send an email, or ask AI to summarize the message.
Zapier, Make and n8n: Popular tools for connecting apps and building automations without writing code. In Zapier an automation is called a Zap.; Example: A Zap that saves every new Facebook lead to a Google Sheet and texts the owner.
Webhook: A way for one app to instantly send data to another the moment something happens.; Example: When a customer pays, the payment app sends a webhook to the website, which marks the order as paid.
API: A standard way for software to talk to other software. Automations use APIs to read and write data in apps.; Example: The website uses Google’s API to add each new lead to a Google Sheet.
NFC: Near-field communication: the tap technology in phones. An NFC card or sticker opens a link when a phone touches it, with no app needed.; Example: A customer taps their phone on a card at the counter and the Google review box opens.
Click-through rate (CTR): The share of people who clicked out of everyone who saw the ad, link or search result.; Example: 3 clicks for every 100 people who saw the ad: a 3% CTR.
Bounce rate: The share of visitors who leave a website without doing anything else. In Google Analytics 4 the opposite, “engagement rate”, is used more.; Example: Most visitors leave the landing page within seconds: the headline or the load time needs work.
UTM parameters: Tags added to the end of a link to show where a visitor came from.; Example: A flyer’s QR code links to …/audit?utm_source=flyer, so the website knows those visits came from the flyer.
Attribution: Deciding which marketing gets credit for a sale when the customer saw several things first.; Example: A customer saw an Instagram ad, then searched on Google, then booked. Which one gets the credit?
KPI: Key performance indicator: a number chosen to show whether a goal is being met.; Example: A salon’s KPIs: bookings per week, rebooking rate and new reviews per month.
North Star metric: The one number that best shows whether the business is succeeding right now. Everything else supports it.; Example: For a new auto shop: booked appointments per week.
Vanity metric: A number that looks good but doesn’t help make decisions or show sales.; Example: 10,000 followers is nice, but calls and bookings tell you more.
Google Analytics (GA4): Google’s free tool that shows who visits a website, where they came from and what they did.; Example: GA4 shows that most bookings come from people who first found the site on Google Maps.