Time-to-market: launch your new products twice as fast

A container arrives at the warehouse on the 3rd of the month, the products are physically available on the 5th, and the first online order comes in on the 28th. In the meantime, nobody has ‘dragged their feet’: everyone has done their job, but always after someone else. This article breaks down this lead time step by step, identifies where time is actually lost, and suggests a reverse schedule that begins before the physical arrival of the goods.

The actual lead time, broken down

Ask your team how long it takes to list a new product online. The answer almost always relates to working time: “twenty minutes per product listing”. That’s correct, but it’s not the point. The lead time you’re interested in is the calendar lead time, from the moment the product is in your possession to the moment it is ready for sale on your channels.

This lead time is broken down into seven stages, usually in the following order:

• Physical receipt— the goods enter stock, are counted and stored. This is the only stage where the date is truly dictated by logistics.
• Obtaining supplier data— technical specifications, dimensions, weight, composition, EAN codes, compliance documents. Often a file received by email, in a format that varies from supplier to supplier.
• Data enrichment— product name, short and long descriptions, structured technical attributes, categorisation, keywords.
• Translation— if you sell in multiple languages, all enriched content must be translated.
• Media— product photos, lifestyle photos, videos, PDF manuals, received from the supplier or produced in-house.
• Validation— marketing proofreading, price checks, regulatory verification depending on the category.
• Publication by channel— e-commerce site, marketplaces, PDF catalogue, Google Merchant Centre feed, B2B portal. Each channel has its own specific requirements.

Try this exercise with your last three product launches: for each stage, note down the actual working time and the calendar time. The difference between the two is the waiting time. In most organisations we see, waiting accounts for the bulk of the delay: a product listing waits for the supplier’s file, then waits for the photographer, then waits for approval from someone on holiday, then waits for the next weekly export to the marketplace.

This is what makes significant time savings achievable without anyone having to work any faster. If your lead time consists of a few hours’ work spread over four weeks of waiting, eliminating the waiting times and running tasks in parallel where possible automatically shortens the timeline — not because we’re working any faster, but because we stop waiting. Conversely, if your process is already stretched to the limit and the steps follow on from one another without any downtime, the gain will be marginal. The scale of the result depends entirely on your starting point: measure it before expecting anything at all.

The five most common bottlenecks

These five bottlenecks recur in almost every distribution catalogue. Each requires a specific countermeasure.

• Incomplete supplier data— you receive a spreadsheet containing the product code, price and a description of thirty characters. Everything else has to be pieced together. Countermeasure: create a standard import template for each supplier, listing the required fields, and reject any data that does not comply upon entry. A supplier who receives specific feedback will correct their export.
• Dependence on a single person— one person knows the structure of the catalogue, the categorisation rules and the specific features of each channel. When they are absent, everything comes to a standstill. Countermeasure: write the rules into the tool rather than relying on someone’s memory — attribute families, permitted values, title templates. Anything that is modelled can be delegated.
• Delayed media— photos arrive after everything else, sometimes several weeks after the goods. Countermeasure: treat media as a separate workflow, initiated as soon as the supplier order is placed. Always request visuals at the time of the purchase order, not when the product goes live. A DAM integrated with the product repository allows files to be linked by EAN code as soon as they arrive, without waiting for the product sheet to be ready.
• Sequential approval— the product sheet passes from marketing to sales, then to management, with each party viewing the content in turn. Countermeasure: approve in parallel on the same content, with clearly defined scopes (marketing validates the copy, sales validates the price; no one validates the whole thing). And set a deadline for tacit approval: if no feedback is received within forty-eight hours, the listing moves forward.
• Re-entry by channel— the same description is re-typed or copied into the website’s back office, the marketplace form, and the printed catalogue file. Countermeasure: a single source and exports by channel. As long as the content exists in multiple places, each additional publication adds to the delay and creates a potential for discrepancies.

Common thread: none of these bottlenecks is resolved by working faster; they are all resolved by changing the order of operations or where the data resides.

The reverse schedule from D-30 to D0

D0 is the target go-on-sale date, not the delivery date. The key question is: what can begin before the product is physically on site? Answer: almost everything relating to the data, provided we have the supplier reference and the technical data sheet.

Only three things must remain sequential. The structuring of attributes must precede editorial enrichment; otherwise, you will be writing based on data that is set to change. Translation must precede multilingual publication, of course. And the launch for sale follows actual stock availability — this is the only physical constraint in the backward planning.

Everything else runs in parallel. The media follow their own schedule from D-30. Translations begin as soon as the source content is finalised, without waiting for approval. Channel mappings are prepared whilst the product listing is being proofread. And there is no reason for the preparation of secondary channels to wait until the main channel is live.

Over a shorter timeframe, the logic remains the same: the key milestone is the availability of the technical factsheet, not that of the colour palette.

Publish incrementally rather than waiting for perfection

Many teams hold back an entire product sheet because the third photo or the Dutch translation is missing. In the meantime, it generates no value anywhere — a costly and rarely conscious trade-off.

The solution is to define a minimum completeness threshold for each channel, rather than a single threshold for perfection. Each channel has its own requirements, and these are not all on the same level:

• Marketplaces— the threshold is imposed from outside. The mandatory attributes per category are a condition for acceptance: if they fall short, the listing is rejected; there is no point in publishing it early.
• E-commerce site— you set the threshold. A proper title, a brief description, a main image, the price and stock levels are generally enough to kick-start sales.
• B2B portal— your business clients are looking for a product reference, a technical specification and availability. Storytelling can wait.
• PDF catalogue or advertising feeds— here the threshold is high, because corrections after publication are impossible or costly.

A completeness score calculated per channel makes this decision-making process automatic: the listing is published on the channels where it meets the threshold, and remains on hold on the others. No one needs to make decisions on a case-by-case basis.

Enrichment then continues on a listing that is already online. This is where automated tools really come into their own:AI-assisted enrichment produces an initial version of the missing descriptions and attributes, which your teams review rather than having to write from scratch. The same logic applies to languages: the machine translation of the catalogue using BYOK enables you to open up a secondary market with accurate content and then refine it, rather than delaying the launch by six weeks.

Measurement: two metrics are enough

Without measurement, a time-to-market project turns into a series of meetings. Two metrics are enough to steer the process, provided they are tracked over time.

• The median time between creation and first publication— in calendar days, calculated based on product listings created the previous month. Use the median rather than the average: just a few exceptional launches are enough to make the average meaningless. Break this down by product family and by supplier; the discrepancies are often telling.
• The proportion of product pages on hold, and why— at any given moment, how many product pages have been created but not published, and which missing attribute is holding them up. It is this ‘by what’ dimension that is decisive: knowing that 240 product pages are blocked is of no use; knowing that 180 are blocked due to the lack of a main image immediately identifies the area that needs addressing.

Record these two figures every month. A falling median lead time indicates that a bottleneck has genuinely been resolved; a lead time that remains static whilst the nature of the blockages changes indicates a problem that has simply been shifted elsewhere.

Frequently asked questions

Can you really halve your lead time?

It depends on the proportion of waiting time in your current lead time. If your work orders spend most of their cycle waiting for a file, a photo or approval, eliminating these delays and running parallel processes where possible yields significant gains, without anyone having to work any faster. If your process already flows seamlessly from one stage to the next without any downtime, the improvement will be minimal. The only way to find out is to measure, for your most recent launches, the difference between the scheduled lead time and the actual working time.

How do you start a product record before you’ve received the product?

Start with the supplier purchase order. It contains the product reference—often the EAN—the purchase price and a description. That’s enough to create the product listing, request the technical documentation and visuals, and begin organising the attributes. The product listing remains in ‘unpublished’ status until receipt is confirmed: it is not visible anywhere, but it is no longer on hold.

Does publishing an incomplete product listing harm search engine optimisation?

There is a risk if you publish an empty listing, with a generic title and no description. This risk becomes negligible if you set a completeness threshold that ensures useful content from the moment it goes live, and then add further details later. A correct listing published three weeks earlier builds up a history and usage signals whilst the ‘perfect’ version is still being proofread.

Do you need a tool to implement this reverse planning?

Reverse planning itself is a matter of organisation, not software. However, three of its key elements become difficult to manage manually once you exceed a few hundred product listings per year: monitoring completeness by channel, staggered publication based on a threshold, and measuring the median lead time. These are precisely the functions that a centralised product repository automates.