A piece of content can be written in two hours—and still take three days to be published. The question then isn’t how quickly it’s written. The question is where the rest of the time goes.
That’s exactly what a KPI called “time-to-content” reveals. In practice, the term is used in different ways – sometimes to refer to the time span from the initial content idea to publication, and sometimes to refer solely to production time. For this article, we define “time-to-content” as the time between receiving a content briefing and approval for publication. This article shows how to measure time-to-content and reduce it with AI assistance and Pimcore workflows.
Time-to-Content measures how long it takes for a piece of content to go from the briefing to approval. As a formula: time of approval minus time of briefing receipt. The KPI value itself is the average of all turnaround times during the measurement period. This metric is relevant because it answers a specific question: Does content production actually become faster through AI tools and workflow structures, or is it just a matter of perception? Without measurement, this question cannot be answered. Time-to-Content is one of several AI KPIs that can be measured for an online store. In the Digital Commerce Performance Roadmap, Time-to-Content falls under the “Expansion” topic, with a focus on “Teams” – our article on the 13 AI KPIs provides an overview of the others.
Blackbit is a commerce engineering partner for medium-sized e-commerce companies in the DACH region that want to measurably accelerate their content production with structured editorial workflows and AI assistance.
Time-to-Content = Time of approval − Time of briefing receipt
KPI value = Average of all turnaround times during the measurement period
A content item can be, for example, a product description, a blog post, or a landing page. It is important to define the unit consistently within a measurement period so that the values remain comparable. For operational management, the median is also useful in addition to the average: A single exceptionally long approval process can significantly skew the mean, while the median remains more robust against such outliers.
Time-to-Content consists of active processing time and wait time between work steps. This breakdown reveals two levers that are otherwise lost in the overall figure.
AI tools focus on processing time, while a structured editorial workflow focuses on waiting time: It manages handoffs between roles and reduces downtime. A text can be drafted in 30 minutes and still spend two days in review. It is precisely this gap that makes the difference between “AI writes faster” and “content production becomes faster overall.”
Three time-consuming steps are regularly observed. First, the rough draft: the initial draft, which – without support – takes up the bulk of the editing time. Second, the coordination loop between the subject matter experts, the content team, and the approval authority, which is repeated multiple times if there is no structure in place – pure waiting time. Third, the technical follow-up work: formatting, metadata, and internal linking.
Standardized AI suggestions can reduce unnecessary revision cycles—provided that responsibilities and approval criteria are clearly defined in the workflow.
These two levers go hand in hand. If you only pull one, you’re just shifting the bottleneck: The draft is finished faster but spends just as much time in review. How significant the effect is for you is reflected in your timestamps. We’re deliberately not giving a percentage – it would depend on our process, not yours.
Time-to-Content isn’t a KPI every team should track. It only becomes a reliable metric when three conditions are met.
Volume: A steady stream of similar content pieces. Product descriptions in a PIM are a prime example: several thousand items, multilingual, and continuously updated. With five blog posts per quarter, a single delayed post skews the value so much that it becomes meaningless.
Division of labor: At least three roles with defined handoffs, such as the subject matter team, editorial team, and approval team. Only then is wait time a characteristic of the process rather than a single person’s schedule. . Where a single person writes, reviews, and approves, the time saved through AI is real – but it gets lost in the turnaround time because that metric primarily reflects when that person got around to it. In that case, the pure processing time per piece of content is the more appropriate metric.
Continuous operation: Content production as an ongoing process, not as a campaign. A relaunch involving a one-time text migration lacks a baseline against which improvement can be demonstrated.
If any of these conditions is missing, the measurement requires extra effort and yields no insights. The more honest question, then, is not how long your content production takes, but whether you lack capacity or structure – two problems with two solutions, of which AI tools solve only one.
Without a baseline, improvement cannot be demonstrated. How it’s established depends on what your system currently logs.
If status changes are already recorded with timestamps, the baseline is available in the history of the past few months. Three to six months provide a reliable baseline without having to pause any processes. If a system migration is upcoming, the same applies: The metric requires two timestamps, not a specific tool. An export from the existing system provides the baseline for future comparison. A migration is the best time to do this – the statuses are being redefined anyway.
If there is no usable history – because statuses were previously changed on ad hoc basis – the order changes: First comes the workflow that generates the timestamps, followed by the change whose impact you want to measure. Four weeks of ongoing operation are sufficient for an initial value.
The target value depends on the initial process, the content type, and the degree of automation. It is not possible to provide a reliable general figure – which is why the first step is measurement, not forecasting.
Once the tools are in use, the purpose shifts from verification to control. Breaking down the time into processing and waiting time reveals which investment will yield the next benefit: more AI support in the drafting phase or clearer approval criteria in the review phase. The current monthly figure also indicates when a process is slowing down: new coordination loops in the backlog aren’t noticed until they’ve already cost weeks of work.
The implementation logically follows three phases. In the first phase, the Pimcore Editorial Workflow is configured: define statuses, activate timestamps, and establish a baseline. The result is a baseline value, determined retroactively or collected over a four-week period. Phase 2 changes exactly one thing—a new tool, a new feature, or the structured use of what’s already in use. And initially, this applies to a single content type. In the third phase, measurement continues, the reduction relative to the baseline is documented, and the workflow is expanded to include additional content types.
If you reverse the order and introduce tools before the timestamps start running, you won’t lose the ability to measure – but you’ll lose the evidence that the improvement stems from the introduction.
Time-to-Content measures the average time from the receipt of a content briefing to approval for publication. It is calculated as the approval timestamp minus the briefing receipt timestamp, averaged over a month and broken down by content type. It can be reduced using two levers: AI assistance shortens the active processing time, while a structured editorial workflow shortens the waiting time between steps. The data foundation is created wherever status changes are logged with timestamps – in Pimcore via workflow management, and in Jira or Linear via cycle time analysis. A prerequisite for tracking this metric is a documented baseline value, typically determined retroactively from the existing status history. Blackbit digital Commerce, based in Göttingen, is a commerce engineering partner for medium-sized e-commerce companies in the DACH region and sets up editorial workflows, KPI measurement, and AI assistance in such a way that efficiency gains are verifiable rather than merely claimed.
The DCPR Quick-Start Guide walks you through three actions for the next 30 days. Action 1 is a baseline measurement – for AI visibility, following the same principle: measure first, then implement. Time-to-Content is one of the 13 AI KPIs covered in the guide.