E-Commerce Consulting.

E-commerce consulting answers whether and how you should sell online. Implementation answers what you build it with. Mix the two and the consulting becomes a sales conversation with an analytical preamble.

An engagement covers four subjects: the product range, the buyers, the channel, and the economics of a single order. Everything else follows from those, including the system question — which in practice is asked first and should be decided last. Anyone who opens with "Shopify or WooCommerce" has skipped the actual decision.

What e-commerce consulting is not: a market study, a competitive analysis in slide form, or a collection of general recommendations. The value comes from your numbers — purchase prices, average basket, shipping costs, return rate, and the working hours genuinely available for catalogue maintenance. Without those numbers you get a document that reads well and decides nothing.

The second difference is that a negative result is allowed. Consulting that may only confirm is not consulting. With us the outcome may be "do not build for now" if the arithmetic says so. It is the same stance we take on prototype validation, where a negative result counts explicitly as a success.

Five decisions determine whether a shop is commercially viable, and none of them is technical. They belong answered in writing before anyone compares systems.

• Product range. Not everything on the shop floor belongs online. What decides is margin, weight and volume, how much explaining the product needs, perishability, the number of variants, and the likelihood of a repeat purchase. A range with a 20 per cent margin and bulky shipping loses money online however good the shop is.
• Buyers. Consumers or businesses, first purchase or repeat purchase, Switzerland only or the EU as well. That answer determines payment methods, price display, legal texts, and the entire structure of your product pages.
• Channel. Own shop, marketplace, social selling, wholesale, or a combination. This question is calculated further down, not felt.
• Shipping and returns policy. Who pays for shipping, above which amount is it free, are returns accepted and who bears them? Switzerland has no statutory right of withdrawal for distance selling. What you grant is a commercial decision, and it belongs in the costing rather than in a template set of terms.
• Maintenance capacity. How many hours a week are genuinely available for catalogue work, order handling and customer enquiries, and who supplies them? This is the most underestimated figure in the whole exercise. More shops die of neglect than of bad technology.

Only once those five points stand does the system question become meaningful. The price bands for building afterwards are published openly on our web design page, which starts at CHF 3,000 and runs to CHF 25,000 depending on scope.

The most important number in e-commerce is the contribution margin per order: the net basket minus cost of goods, payment fee, shipping and packaging, and the attributable share of returns. Only that amount is available for advertising, fixed costs and profit — and it regularly comes out smaller than the gross margin suggests.

Here is the calculation with realistic Swiss values. A basket of CHF 80 contains roughly CHF 74 of net revenue at 8.1 per cent VAT. From that, deduct 45 per cent cost of goods, which is CHF 33.30. Deduct about 2 per cent in payment fees, CHF 1.60. Deduct CHF 9 for shipping and packaging. Deduct CHF 3 of returns cost, calculated as a 15 per cent return rate multiplied by roughly CHF 20 of handling per return. Around CHF 27 per order remains.

Now comes the part that sinks most shop plans. If a click in paid search costs CHF 2 and one visitor in fifty buys — a conversion rate of 2 per cent, an ordinary figure in Swiss retail — then an advertised order costs CHF 100. Against CHF 27 of contribution margin that is a loss of about CHF 73 per order. For paid advertising to break even at this basket size, the conversion rate would have to sit near 7 per cent. Shops practically never reach that.

Three sober conclusions follow. First: at small baskets a shop only works on cheap traffic — organic search, existing customers, repeat purchases, referrals. Second: raise the basket and the picture changes. At CHF 200 with the same cost structure the contribution margin is roughly CHF 86, and at a 2 per cent conversion rate a click may cost up to about CHF 1.70 before the order goes negative. Third: anything that lifts the conversion rate acts directly on this calculation. The groundwork is set out in our article on conversion rate optimisation in Switzerland, and none of these figures is reliable without clean measurement, as our page on conversion tracking explains.

In an engagement we build this table with your numbers, not with example values. It is unspectacular, and it is the reason some projects are finished after two hours.

The channel question turns on three quantities: the fixed costs you can carry, the commission your margin tolerates, and whether you need the customer relationship. The pure cost calculation is quickly done — and it is only half the answer.

The cost threshold itself is straightforward. An own shop with around CHF 300 of monthly fixed costs beats a marketplace charging 12 per cent on a CHF 80 basket from roughly 31 orders a month. Include the one-off build at CHF 12,000 over three years and the threshold moves to about 66. In an engagement we run the same threshold on your figures — your basket, your commission rate, your actual fixed costs. It shifts considerably: at a basket of CHF 200 instead of CHF 80 the same commission carries far more per order, and the cost threshold falls from roughly 31 to roughly 13 orders a month.

The other half of the answer is strategic. A marketplace delivers reach on day one, but it stands you directly beside comparable offers and keeps the customer relationship. Swiss online retail is shaped accordingly: among the 30 highest-turnover online shops in the country, 13 were already marketplaces in 2025, against a total volume of CHF 15.8 billion. Selling there means selling in someone else's shop window, with everything that brings in reach and costs in control.

In many cases the sequence is the right answer: test on a marketplace whether the range finds demand, then build your own shop once repeat purchases and direct enquiries prove a customer relationship exists. For ranges that do not exist yet, an even smaller test is better — a single order page with a payment connection, as described on our page about prototype validation.

The make-or-buy question turns on three criteria: the share of your processes that departs from the standard, the total cost over the useful life, and the cost of getting out later. The purchase price is the worst of these criteria and still usually the only one compared.

• Deviation share. List the ten operations your shop must support. Eight covered by the standard and two by extensions: rent. Four that only work with custom development: build. It is crude, and it settles the decision better than any feature matrix. The principle behind it is explained in our glossary entry on make or buy.
• Total cost over the useful life. Platform or hosting fees, app subscriptions, transaction fees, maintenance, customisation and internal effort across five years — set against the one-off build plus maintenance for an own development. What belongs in that sum is set out in our entry on total cost of ownership.
• Exit cost. What does moving in three years cost? Hosted systems export data readily; the design and the app-driven functionality do not travel. Almost nobody asks this before signing, and almost everybody asks it after two years.

We apply the same logic to other standard software. The full five-year calculation is set out on CRM: buy or build. If the arithmetic points to an own development, custom software in Switzerland describes the process and who owns the result.

An engagement is three meetings and one document. The scope is deliberately small, because the questions are important but finite.

In the first meeting we take the starting position: range, purchase prices, expected basket, audience, existing channels, existing systems, and the real maintenance capacity. Between meetings we work through the contribution margin per order, compare the channel options against your numbers, and test the system question on the three criteria above. In the second meeting we present the result and argue it out, explicitly including arguments against our own recommendation. The third meeting is a short close with the next steps.

You receive a document in five parts: the per-order costing, the channel decision with its reasoning, the system recommendation, a list of the points that must be settled before any build — shipping rules, payment methods, tax rates, legal texts, product data — and an estimate of the internal effort in day-to-day operation. The document is written so that you can obtain comparable quotes from several agencies with it. That is deliberate.

Consulting engagements are billed by effort, and we name the range after a short preliminary conversation. Full disclosure belongs here: we also build shops. Every consulting result therefore states which recommendation would earn us follow-on work and which would not, so you can read it rather than guess at it. Where we can point to delivered e-commerce work, we do: JUCAN Tailored Suits received a website, a Shopify store and a complete brand identity from us.

In four situations consulting is wasted money, and we say so in the preliminary conversation. An engagement with nothing left to decide is an expensive set of minutes.

• The decision is made and reasoned. If range, channel, payment methods and shipping policy are fixed and the numbers behind them have been checked, you need a build, not advice. Go straight to the implementation conversation.
• The shop already runs and the problem is measurable. If orders arrive but the basket is too small, or the checkout produces drop-offs, that is not a strategy question. Work on measurement and conversion instead of on a concept.
• You sell a single product in small quantities. At five orders a month, consulting costs more than the mistake it is meant to prevent. Take a marketplace or a simple order page and test demand directly.
• The numbers are not available. Without purchase prices, shipping costs and a defensible assumption about the basket, the contribution margin cannot be calculated. Get those numbers first — they are the basis of every later decision anyway.

If a question survives those four points, an engagement makes sense. If none survives, you have advised yourself — and that is the best outcome this page can have.