eCommerce

    E-commerce consultant: why selling hours is no longer enough

    Francesco Luigi Ziccheddu · November 18, 2026 · 7 min
    Online shop desk at night, with a screen showing a sales dashboard glowing in the dark

    The average consultant sells hours. It is an honest way to work, but it measures the wrong thing. An online seller's problem does not arrive during office hours: it arrives at three in the morning on a Saturday, over the Black Friday weekend, or in the second half of August, when half of Italy is on holiday and a wrong price list keeps on selling. My thesis is simple. An e-commerce consultant only makes sense if they also cover the hours when they are not there, and today the serious way to do that is to put AI agents that keep watch alongside the people, and a person who decides.

    Why selling hours is the wrong model for an online shop

    An hourly contract works when the job has a beginning and an end. You write a plan, reorganise a catalogue, set a pricing strategy, and at some point the work is delivered. The trouble is that an online shop is not a project: it is a system that stays on twenty-four hours a day, seven days a week, and can break at any moment. A product listing gets suppressed, a price drops below cost after a feed update, a stock line runs out while the advertising keeps spending. None of these events waits for Monday morning.

    Whoever buys hours buys presence, and a professional's presence has a timetable. The client, however, does not have a timetable problem, they have a continuity problem. When the consultant can only be reached on weekdays, the owner ends up guaranteeing continuity personally, with the phone on the bedside table and an eye on marketplace notifications even in mid-August. It is a service that covers the risk on paper and leaves it exactly where it was in practice.

    There is also a second effect, less visible. If the fee depends on hours, the consultant has no incentive to build a system that makes those hours unnecessary. Every automatic check that removes manual work shrinks the invoice. It is not bad faith, it is the shape of the contract itself. And the client ends up paying, month after month, for checks that a machine could run in a few seconds.

    What really happens at night, at weekends and in August

    Consider three scenes that anyone selling online will recognise. The first: on Saturday night a listing is suspended because of a missing field. Until Monday, sales of that product are zero, and if it is one of the ten products that make up half the turnover, the damage is counted in days of revenue, not minutes. The second: in Black Friday week, which this year falls on Friday 27 November with Cyber Monday on the 30th, a competitor cuts their price on Thursday evening and your product loses its best visibility without anyone noticing. The third: in August the office is closed, the consultant is on holiday, and the supplier announces a delay that will run the best seller out of stock just as orders pick up again.

    Phone on a bedside table showing a marketplace notification in a dark bedroom

    In all three scenes the value is not in knowing how to run the analysis. It lies in noticing in time. And noticing in time is a job of watching, not of reasoning: looking every day at the same accounts, the same prices, the same stock levels, and flagging when something drifts from normal. It is the kind of work in which a person is expensive and intermittent, and an automatic system is cheap and tireless.

    What changes with AI agents

    Until recently this idea stayed theoretical, because building a continuous monitoring system cost about as much as a bespoke management system. Today AI agents have lowered the threshold. You can see it on the shopper's side too: this week BigCommerce announced a Shopping Agent, a conversational assistant that stores can customise and that helps visitors search for products, compare them and reach the basket, rolling out right now ahead of the gift season. Like it or not, buyers increasingly talk to software that never sleeps. If the demand side works around the clock, the supply side cannot rely on a watch that runs from nine to six.

    I want to be clear, though, about what an agent is not. It is not a consultant in miniature, and it does not make the decisions that matter: whether to lower a price, whether to accept a promotion, whether to pause advertising on a product that is running out. An agent observes, compares against what is normal, prepares the work and raises the alert. The decision stays with a person, who arrives in the morning with the problem already described and the options already prepared, instead of starting from scratch. The consultant's time shifts from finding the problem to solving it, and that is the change worth paying for.

    How to organise a watch that never sleeps

    Method matters more than tools. The first question is not which agent to buy, but which three or four events would really hurt if they stayed invisible for forty-eight hours. Suspended listings, price gaps, stock falling below a threshold, anomalies in advertising cost. You start from the process, not from the tool, and you write down precisely what the system can do on its own and what requires approval.

    Empty office in August, with a tidy desk and a screen still switched on

    That second point is what separates a serious setup from a risky experiment. In sensitive processes the AI prepares the work and a person checks it. An agent can flag an out-of-line price and propose the correction, but the change to the price list goes through whoever answers for it. The same applies to customer service: drafted replies, routed requests, ready summaries, with the last word always going to a person. It is the logic BusinessLabs applies to its AI and automation work, and the one behind Selleee, which integrates AI agents into the work of those who sell on Amazon. Automation that works is automation that reduces checking work without removing responsibility.

    The third element is rhythm. A watch that raises too many alerts gets ignored, and one that raises too few is useless. It is better to start with a few rules, see over a month which alerts proved useful, and drop the rest. A good monitoring system is recognisable by the fact that every message it sends is one that somebody wants to read.

    How to choose an e-commerce consultant today

    Anyone choosing a consultant can ask a few very concrete questions. What happens if a listing is suspended on a Saturday night? Who notices, and after how long? Which checks are automatic and which depend on a person being available? What is delivered at the end of the month: hours worked, or a system that keeps running even when nobody is looking at it? And above all, who decides when a decision is needed, and with what information in front of them.

    A consultant who can answer these questions is not selling hours, even if hours are involved. They are selling continuity, which is the only thing the client really needs. Someone who answers that on Monday morning they will run a full check is describing a model built around their own calendar, not the shop's.

    What this means for an Italian business

    For an Italian company that sells online, and that earns a significant part of its year in the weeks of Black Friday and Christmas, the risk is not the mistake itself but the mistake that lasts. A problem found after an hour costs little. The same problem found after three days, in a peak weekend, can wipe out the margin of an entire campaign.

    My position is this. The consultant who only sells hours will find it harder and harder to justify their fee, because the most repetitive part of their work can be done better by a system that checks every day. What will remain, and be worth more, is judgement: knowing what to do when the system raises an alert. Whoever builds their service around judgement and continuity, rather than around the stopwatch, will be the right partner for online sellers. The client never had a problem of hours. They have a problem of nights, weekends and August, and deserve someone who looks after it then as well.

    Francesco Luigi Ziccheddu
    Author
    Francesco Luigi Ziccheddu
    CEO & Co-Founder

    CEO & Co-Founder of BusinessLabs. I have attended major international trade shows since 2015 and share here experiences and insights for online sellers.

    LinkedIn

    Talk to our Amazon agency

    Book a free consultation