People keep saying you need a content team and daily posts before GEO can work. That is only half right. Content does need to keep flowing, but treating GEO as a volume problem is the mistake. AI does not care how much you publish; it cares whether clear, trustworthy material exists to quote. So GEO is a mechanism problem, not a headcount problem: turn the questions clients actually ask into searchable answers, publish on a rhythm one person can keep, and check monthly whether AI cites you. That is a freight forwarder GEO setup you can run alone.
What to write: your clients' questions are the topic list
Most people get stuck on "what do I write about," yet the answer sits in your account managers' chat logs. The questions clients repeat are your best topics: how many days a route takes, whether a quote includes customs clearance, whether you can handle dangerous goods. Copy those questions, de-duplicate them, sort them by trade lane and cargo type, and twenty will carry you through a month.
This content has an advantage: what clients ask, AI also gets asked, so matching answers get quoted more often. Industry commentary sits too far from any specific question for AI to use. A survey of cross-border sellers points at the same thing (Marketplace Pulse 2026 Seller Index): 83.4% use AI already, yet 25.4% say it has produced no measurable result. The difference is not whether you use AI, but whether your content sits inside a mechanism you can verify.
How often to write: a rhythm one person can keep
You do not need to post daily. The smallest workable rhythm is two client Q&As a week, and an hour or two covers it. What matters is a steady week, not a heroic one: two posts every week beats ten in one burst followed by a silent month. Ongoing updates are what keep search engines and platform algorithms recommending you; stop and start, and the authority you built quietly cools.
How to check it is working: ask the AI once a month
How do you know your content is doing anything? Take a few questions clients ask often and put them to AI tools such as Doubao, Kimi and Yuanbao once a month, then note whether your brand shows up. If it does, your sources are working. If it never does, look back: is the question uncovered, or is the page structured in a way AI cannot read? This check takes under twenty minutes and saves you from writing for half a year without knowing what lands.
A Shenzhen freight forwarder's real numbers
We worked with a cross-border freight forwarder in Shenzhen. Five account managers handled over 200 inquiries a day, each quote took 30 minutes, and conversion was 3%. We organised their quoting logic and common questions into a knowledge base and added an AI quoting assistant. Quote time dropped to 2 minutes, conversion rose to 8% (measured on that client project in 2026), and service cost fell by roughly 70%.
The same set of client questions, organised once and reused many times — that is the compounding value of a question bank. It does not depend on inspiration, only on writing down what you have already said.
The rhythm on one sheet
- Every week (about 2 hours): write 2 real client Q&As, pulled from the account team's chat logs.
- Every month (about 1 hour): check the basics on your website and every platform account, so the name, address, phone and service scope match.
- Every quarter (about 1 hour): review new questions clients raised, add them to the bank, drop the outdated ones.
That averages four to six hours a month — far cheaper than a dedicated content role.
Two traps to avoid
First, handing the writing to someone who does not understand the business. The result is hollow, with no real cases, so AI will not quote it and clients will not read it.
Second, buying a one-click GEO tool and waiting. The tool can publish, but what to write and whether it is accurate remain human calls. Tools handle the hands; the mechanism handles the thinking.
Frequently asked questions
Can one person really run GEO without a team?
Yes, if you lock three things down: what to write, how often to write, and how to verify. The question bank comes from clients, the rhythm is two hours a week, the check is once a month.
Is two posts a week too few?
For a small forwarder it is enough. GEO rewards consistency and accuracy, not volume. Two posts a week for three months beats thirty posts in one month and then silence.
Should I write client Q&As or industry commentary first?
Client Q&As first. Most of what AI gets asked is a specific question, and specific answers are easier to quote. Commentary usually brings fewer inquiries.
Should I write it myself or hand it to an agency?
If the industry knowledge lives only in the owner's head and the team has no time to organise it, start with an agency that has an industry knowledge base; it usually takes one to two months to get the question bank and rhythm running. To do it yourself, start today by noting twenty real questions.
Shanghai Naili has run freight-forwarder marketing and content for 10 years, and we have watched too many companies treat GEO as a one-off purchase. It is a small mechanism you run daily: questions come from clients, answers go back to clients. The real gap is never who writes the most, but who says it clearly enough that AI is willing to say it again for them.