Process Automation

    Process automation with measurable results

    With process automation from datadice, a recurring process runs through your existing systems, such as DATEV or your ERP. A person only sees it when a decision is due. Before we connect the first process, we agree with you on a metric, for example the lead time per incoming invoice. Then we measure the baseline.

    In McKinsey's survey “The state of AI in 2026” with 1,719 participants, 80 percent of those who use AI at work report higher personal productivity. 37 percent of respondents see an impact on their company's EBIT.

    Goal:Automate routine work

    Last reviewed: · reviewed by Thomas Knorr, Managing Director datadice

    Incoming invoice R-4711Checking
    Purchase order 2048
    Delivery note
    Amount €7,200
    Waiting for your approvalApprove

    Metric: lead time per incoming invoice

    Example with fictitious data

    At a glance

    What
    datadice automates processes in which data is captured several times and reconciled by hand: invoice checking, order entry, purchasing.
    For whom
    Companies whose accounting or order processing does the same reconciliations by hand every month.
    Outcome
    An automated process across your systems, documented business rules, monitoring and a metric with baseline and target value.
    Getting started
    Initial consultation, then a Strategy workshop with process measurement

    Funding:Small companies in Bavaria receive a grant of up to 50% from the Digitalbonus for the introduction. We prepare the application with you free of charge.

    Go to the Digitalbonus

    A typical starting point

    In your accounting department, incoming invoices are captured, matched against the purchase order and delivery note and forwarded for approval. Similar routines exist in order entry and purchasing. This work takes up time every month, and errors can mean that early-payment discount deadlines are missed.

    Our answer

    Your specialists work on tasks that require knowledge and experience

    Automation takes over duplicate data entry and matching. Control over rules, exceptions and approvals stays within your company.

    What we stand for
    We are convinced: specialists belong on tasks that require their knowledge and experience. Entering data several times and reconciling it by hand is not one of them. Automation must therefore improve how systems work together, noticeably relieve employees and create economic benefit. Control over rules, exceptions and approvals stays within the company.

    Definition

    What is automated invoice processing?

    Automated invoice processing reads an incoming invoice, matches it against the purchase order and goods receipt and presents it ready for approval. At datadice, a person always gives the approval itself.

    Many programs already handle the reading well, DATEV included. The costly work comes afterwards. Whether the price matches the order, whether the full quantity has been delivered and who in purchasing placed the order: accounting and purchasing often clarify this by email, phone and spreadsheet, invoice by invoice.

    This is exactly the part datadice connects. The process fetches the data from your ERP, checks it against your rules and notes why it passes or holds an invoice. Routine cases then sit in the inbox ready for approval; deviations go to the responsible person with a reason.

    For your finance team

    Is time saved already a saving?

    No. Freed-up working time only turns into money when your team uses it for something else or a position does not need to be refilled. That is why datadice calculates three effects separately: relief, avoided costs and additional revenue.

    A worked example with assumed values: 1,200 incoming invoices a month, six minutes of matching per invoice, that is 120 hours. If the automated process prepares 80 percent of the invoices ready for approval, 24 hours of checking remain. At first, the 96 freed-up hours are only relief.

    Avoided costs arise when no additional accounting staff is needed or missed early-payment discount deadlines disappear. On the other side are the costs of introduction, internal effort, licences and operation. The real calculation uses your measured values from the process measurement.

    Worked example with assumed values.

    For your department

    What happens with partial deliveries, price deviations and other special cases?

    The process holds special cases and sends them with a reason to the person who already decides on them today.

    In daily work, the exceptions are what count: the partial delivery that is only complete next week, the supplier with the new discount, the invoice without an order number. datadice records these cases with your team during the process measurement, using real cases from recent weeks. They become the acceptance cases.

    Only when the process also recognises the difficult cases correctly and passes them on cleanly is it accepted. Until then it runs in parallel with the existing process, and your team compares the results. Whoever maintains the rules later learns how in the datadice Academy.

    Architecture

    How datadice automates a process

    In the connected process, your systems match invoice, purchase order and delivery note, so only the cases that need a decision reach your team.

    Step 1 of 5

    01

    Capture

    Together with your team, we capture processes, data flows and exceptions.

    Example: checking incoming invoice R-4711Step 1 of 5

    Capture

    Process: Checking incoming invoices

    Price deviationpartial delivery

    Outcome

    What you receive

    You receive an automated cross-system process with documented business rules, agreed system connections and monitoring. One example is invoice checking: the process reads the invoice, matches it and prepares the approval.

    • Automated cross-system process
    • Documented business rules
    • Agreed system connections
    • Process monitoring
    • Exceptions routed to the responsible person with a reason
    • Metric with baseline and target value
    • Enabling your team to monitor the process and maintain the rules

    The first step

    The first step is the process measurement

    The datadice process measurement records a clearly defined process and provides the economic basis for your decision. We measure volumes, processing times and exceptions in your operating data and separate measured values from assumptions.

    You receive a documented process record, the measured baseline, a cost-benefit calculation, the recommended implementation scope and the acceptance criteria. The fixed-price offer for the implementation is based on these.

    Comparison

    Standard software, custom development or datadice?

    The table shows at which point in a process each approach comes in.

    Starting point

    Standard software
    the product's standard process
    Custom development
    the project's specification
    datadice
    a clearly defined process with a metric

    Connection to existing systems

    Standard software
    via the vendor's existing interfaces
    Custom development
    depending on the contract
    datadice
    your systems, such as DATEV or your ERP

    Your own business rules

    Standard software
    within the configuration options
    Custom development
    freely implementable
    datadice
    documented business rules, exceptions routed to the responsible person with a reason

    Quality assurance

    Standard software
    by the vendor
    Custom development
    as agreed in the project
    datadice
    parallel operation and comparison against a jointly agreed metric

    Operation and further development

    Standard software
    with the vendor's releases
    Custom development
    by separate agreement
    datadice
    process monitoring to the agreed extent

    Enabling your team

    Standard software
    product training
    Custom development
    as required
    datadice
    your team maintains the rules and monitors the process

    Application

    What is straight-through processing?

    Straight-through processing is the share of cases that run through without human intervention. At datadice it extends up to the prepared approval; the approval stays with a person.

    The rate is one of several metrics for an automation project. For incoming invoices, lead time usually matters most, because early-payment discount deadlines depend on it. In order entry, the error rate matters more, because an incorrectly captured order causes work later in production and shipping.

    If a process waves through 95 of 100 invoices and nobody understands the remaining five any more, nothing is gained. That is why every exception comes with a reason your team can read without asking back.

    Frequently asked questions

    An objection we often hear

    „DATEV already reads our documents.“
    Good, then the reading is done. datadice automates the part after it: matching against the purchase order and delivery note and preparing the approval. The approval stays with a person.
    How much does process automation cost?

    The entry point is the Strategy workshop at €3,900 net. What the implementation costs depends on the process; we clarify the scope in the initial consultation.

    Who decides on exceptions?

    The responsible person in your team. They receive every exception with a reason.

    Which processes are suitable?

    Processes that run the same way in larger numbers every month and need data from several systems: incoming invoices, order entry, purchase orders.

    Getting started

    Which process takes up the most time in your company every month?

    In the free initial consultation, Thomas Knorr shows the calculation using invoice checking as an example. Afterwards you receive a short playbook with our assessment and the next step.

    Starting point:Strategy workshop with process measurement