Category: Blog

  • Generative Engine Optimization (GEO): How to Get Mentioned in ChatGPT Answers

    Generative Engine Optimization (GEO): How to Get Mentioned in ChatGPT Answers

    Ask an AI assistant “which company should we use for …” and it rarely answers with a list of links. It answers with names. Those answers increasingly shape shortlists, for consumers and for procurement teams alike. Your company is either in them or it isn’t.

    Getting into those answers is called generative engine optimization (GEO), sometimes answer engine optimization (AEO). It is the work of being understood, cited and recommended by ChatGPT, Perplexity, Claude, Gemini and Google’s AI results.

    Why it matters now

    • Buyers research inside AI assistants before they visit a website, especially for B2B services and software.
    • AI answers sit above the links on many commercial searches, so fewer people click through to the underlying pages.
    • Ranking and being cited are different. A page that ranks first on Google is not automatically the source an AI quotes, and the reverse is also true.

    Strong SEO is still the foundation. But a company can rank well and still be missing from the answers its buyers read.

    What AI engines look for

    Clear, citable statements. Specific claims an engine can lift and attribute. “We help companies grow” is never quoted. A precise description of what you do, for whom, and how, can be.

    Trusted mentions. Coverage in industry publications, directories, podcasts, YouTube, Reddit and other places AI models learn from and search.

    Entity clarity. A consistent picture of who your company is: name, services, leadership, locations served, all described the same way across your site, profiles and listings, backed by structured data.

    The seven things that move GEO

    1. Rewrite key pages for citation. Replace vague claims with specific ones, add clear definitions, numbered steps and plain-language summaries.
    2. Add proper schema markup. Organization, Service, FAQ, Article and HowTo, nested correctly and describing content that is really on the page.
    3. Make your entity consistent. Same name, description and details on your site, LinkedIn, Google Business Profile, directories and press.
    4. Earn mentions where AI looks. Industry publications, podcasts, YouTube, communities and comparison pages in your category.
    5. Publish answer-shaped content. How-to guides, comparisons, definitions and “best for” pages that match the questions buyers ask.
    6. Put FAQs on service pages. Short, specific questions and answers are easy for engines to extract.
    7. Track AI visibility. Measure how often your brand and competitors appear for the prompts that matter, across each engine, and work from the gaps.

    What doesn’t work

    • Keyword stuffing. Engines want natural language and clear claims.
    • Unedited AI content. Generic text produced at volume is rarely cited by AI. Specifics and original insight are.
    • Schema without substance. Markup on a thin page does not fool anyone.
    • Paid links on low-quality sites. They don’t move AI visibility and can damage your search standing.

    A realistic timeline

    • Weeks 1–2: visibility audit across engines and prompts, gap analysis, plan.
    • Weeks 2–8: schema, page rewrites, FAQ blocks, entity cleanup, first outreach.
    • Months 2–3: first citations on lower-competition prompts. Engines that search the live web tend to pick up changes first.
    • Months 3–6 and beyond: citations compound as mentions and content build up.

    Like SEO, GEO compounds. Companies that start early hold their position before categories get crowded.

    Do it in-house or with a partner?

    The tactics are documented, so an in-house team can do it. What is hard is the volume (rewrites, schema across a whole site, FAQs, outreach and tracking) and the editorial judgment to write content AI engines actually quote. Many companies run GEO with a partner next to their own marketing team.

    We run SEO and GEO programs for companies of every size. See our marketing and search work, or get a free AI audit to see where your company stands.

  • AI Chat Agents vs Live Chat: What Actually Works

    AI Chat Agents vs Live Chat: What Actually Works

    Every company with a website eventually asks the same question: should a person answer our website chat, or should AI?

    The honest answer for most companies is both, in the right order. Here is how to decide.

    What each one is good at

    AI chat agents

    • Coverage. They answer at 2am, on weekends and during spikes, without a queue.
    • Consistency. The same correct answer about your products, policies and hours, every time.
    • Speed. Replies in seconds, which matters most for lead capture.
    • Volume. They handle a hundred conversations as easily as one.

    Live chat with people

    • Judgment. Unusual requests, upset customers and anything with nuance.
    • Complex sales. High-value deals where trust and conversation matter.
    • Exceptions. Refunds, escalations and decisions outside policy.

    Where companies go wrong

    All human, limited hours. Chat is staffed nine to five, so most visitors after hours meet an offline form and leave.

    All AI, no exit. A bot that cannot hand off traps frustrated customers in a loop. That costs more trust than having no chat at all.

    A generic bot. An agent that was never properly trained on your business gives vague or wrong answers and makes the company look careless.

    The model that works: AI first, people on call

    1. The AI agent answers every conversation first. It is trained on your real products, policies, pricing rules and FAQs.
    2. It captures and qualifies leads and books meetings straight into your calendar and CRM.
    3. It hands off to a person when the customer asks, when sentiment drops, or when the question is outside its rules, with the full conversation attached.
    4. People review the transcripts regularly and fix gaps, so the agent gets better every month.

    For a smaller company, the handoff can go to email or a call-back. For a larger support operation, it routes into your live chat or help desk queue, by team, language or product.

    What to measure

    • Resolution rate: the share of conversations the agent resolves without a handoff.
    • Lead capture: how many chats turn into contacts and booked meetings.
    • Handoff quality: whether people get the context they need without asking again.
    • Customer satisfaction on both AI and human conversations.

    Choosing for your company

    If you have steady web traffic and a small team, start with an AI agent and a clean handoff to email or phone. If you run a support team, put the agent in front of your existing live chat and let it take the repeat questions. If you operate several brands or regions, plan the routing rules and knowledge base first; that is where most of the value sits.

    We build and run AI chat agents with human handoff and weekly quality review. See our customer service work, or get a free AI audit and we will tell you what fits your company.

  • Where AI Gives Your Team Hours Back Every Week: A Practical Rollout Plan

    Where AI Gives Your Team Hours Back Every Week: A Practical Rollout Plan

    Most companies don’t have a productivity problem. They have a “people doing work a machine should be doing” problem.

    You have probably read a hundred articles about AI. Most are either predictions about how it will transform everything, or reviews of the latest tool. Neither gets your team’s week back.

    This guide is the practical version: where AI actually takes work off people, the order to roll it out so you don’t waste money, and where to keep a person firmly in charge. It applies whether you have twelve employees or twelve hundred. The volume changes. The pattern doesn’t.

    Where the hours actually go

    When we run an audit, the same categories of work show up in almost every company:

    • Answering the same customer questions by phone, chat and email, every week
    • Slow follow-up on inbound leads, because nobody owns the first response
    • Writing routine emails, proposals and content from scratch
    • Copying data between tools that should be connected and aren’t
    • Categorizing transactions and chasing invoices
    • Scheduling, rescheduling and reminders
    • Pulling numbers into a weekly report by hand
    • Screening applicants during hiring waves

    None of this grows the business. It keeps it running. AI can take most of the volume. Your people keep the judgment.

    The highest-payback moves, roughly in order

    1. Answer every inbound conversation

    A chat agent on your website and a voice agent on your phones answer the repeat questions, capture leads and book calls around the clock. For companies with real web traffic or call volume, this is usually the fastest win, because it recovers leads you are currently losing to voicemail and bounce. How we build voice and chat agents.

    2. Respond to every lead within minutes

    Speed to lead is one of the most reliable levers in sales. An agent that replies to every new inquiry right away, asks a qualifying question and offers a calendar link fixes slow follow-up in a single workflow.

    3. Connect the tools your team copies between

    Lead comes in, CRM updates, email goes out, meeting is booked, finance is notified. Workflow automation on n8n, Make or custom code removes the manual handoffs. In larger companies this is where the biggest hidden hours sit. See our automation work.

    4. Triage support tickets

    For teams with real ticket volume, AI that tags, routes and drafts replies lets agents approve instead of type. Simple cases close themselves; difficult ones reach a person faster.

    5. Automate the paperwork

    Quotes, contracts, invoices and onboarding documents drafted from your own data, sent for signature and tracked automatically.

    6. Clean up the books

    Automatic categorization, receipt capture and reconciliation rules in QuickBooks or Xero, with a bookkeeper reviewing the month-end close.

    7. Fill the pipeline

    For B2B companies, an AI lead engine plus outreach and dedicated sales agents replaces manual prospecting and fills calendars with qualified calls. Sales and lead generation.

    8. Produce content without writing it yourself

    Articles, social posts and video produced with AI and edited by people, built to rank on Google and get cited by AI assistants.

    9. Replace the Monday spreadsheet

    A live dashboard across your tools, plus a short AI-written summary of what changed each week.

    10. Train your people

    Many companies already pay for AI tools that few staff use. Hands-on training on real tasks is often the cheapest win on this list. AI training for your teams.

    The order to roll it out

    Doing everything at once is a mess. A sequence that works:

    1. Month 1: stop the leaks. Answer every call and chat, respond to every lead fast, start asking for reviews.
    2. Month 2: build the pipeline. Lead engine, outreach and sales agents if you sell B2B.
    3. Month 3: fix the back office. Automate the five most painful manual workflows and the paperwork.
    4. Month 4 onward: content, reporting and training. Compounding work that pays back over quarters.

    Larger companies run the same sequence department by department, with a pilot team first and a wider rollout once the numbers hold.

    What not to automate

    Final calls on disputes and refunds. AI should triage, not decide.

    Hiring decisions. AI can screen. People choose.

    Anything regulated. Medical, legal, credit and individual financial decisions need a qualified person. AI can support them, not replace them.

    Watch out for

    • “AI” that is really an outsourced team with a chatbot wrapper. Ask exactly which tools run the work and who checks it.
    • Outreach without deliverability discipline. Sending cold email from your main domain on day one burns your reputation.
    • Unedited AI content. Mass-produced articles are exactly what search engines filter out.
    • Pilots with no owner. If nobody is responsible for monitoring and improving an agent after launch, it quietly degrades.

    Where to start

    The fastest way to find out which of these moves pays back first for your company is the free AI audit. You answer a few questions about your business, team and tools, and get a written audit by email with your top opportunities, a 30-day plan and honest notes on what to skip.

    Get your free AI audit, or book a 30-minute call with a senior consultant.

Get my free AI audit →