Customer Service Email Automation: How myphotobook Automates 83% of Emails

Every CX leader knows the pattern: peak season arrives, campaigns perform, and the support inbox fills faster than any team can type. The strategic question is not whether to automate, but what automation actually delivers on the P&L. The case of myphotobook, one of Europe’s leading providers of personalised photo products, gives a concrete answer: 83% of all service emails are now resolved automatically through customer service email automation with OMQ, saving around €408,000 per year at consistent service quality. This case study breaks down the setup, the numbers and what other organisations can take from it.

The case study at a glance

All figures come from the joint success story on OMR Reviews (June 2026).

The starting point: millions of memories, one email mountain

Personalised photo books, calendars and wall art come with questions: Which paper quality is right? Will the pictures print sharply? Can an order still be changed? Will the photo book arrive in time for the holidays? myphotobook answers such questions thousands of times a year, in several languages, for multiple brands, around the clock.

What works in quiet months becomes a stress test in peak season. Before Christmas, enquiry volume spikes and customer expectations rise with it: anyone ordering a gift wants certainty that it will arrive on time. At the same time, the very same team is expected to resolve complex complaints and emotional cases. Without technological support, a traditional service team hits its limits exactly when quality matters most, and that shows up in CSAT and repeat purchase rates.

Added to that is a pattern every service organisation recognises: the same topics keep recurring. Where is my order? Can I still change my photo book? Which file formats are accepted? As long as the team answers these standard questions manually, no capacity remains for the cases that genuinely need human judgement.

The strategy: one knowledge base, every channel

Many organisations automate channels in isolation: a chatbot here, an FAQ section there. The result is knowledge silos, gaps and contradictory answers, and any efficiency gain evaporates in maintenance overhead. myphotobook took a different route from the start and built an overarching automation strategy together with OMQ.

The centrepiece is the central knowledge base: it powers every channel simultaneously, whether email (OMQ Reply), chat (OMQ Chatbot), help page (OMQ Help) or contact form (OMQ Contact). When a team member updates a piece of information, it is instantly live everywhere. No manual re-maintenance, no contradictory answers across channels. What began with the chatbot has grown over the years into a connected service ecosystem, and its current culmination is email automation with OMQ Reply.

OMQ Reply in detail: two modes for maximum control

OMQ Reply is built on a multi-stage LLM pipeline: the AI email agent understands incoming messages, classifies the context and formulates precise answers. Depending on the enquiry type, two modes are available:

Draft mode: the AI proposes, the team decides

The AI analyses incoming emails and creates a ready-to-send answer draft directly in the ticket, based on the service knowledge. Employees review it briefly and send it with one click. That saves time without giving up control, ideal for sensitive topics or the roll-in phase.

Auto-reply mode: standard enquiries run fully automatically

Recurring enquiries such as order status or file formats are answered fully automatically, around the clock, with a built-in confidence safeguard: only when the system is certain about the answer is the email sent automatically. Everything else lands with the team, which only steps in when it is genuinely needed.

At myphotobook, integration into the existing Zammad ticket system took just a few minutes. A months-long IT project is not a prerequisite for starting with automated customer communication, which matters for any evaluation where time-to-value is a decision criterion.

The results: 83% automation, €408,000 in savings

What customer service email automation delivers is best shown by the numbers:

An 83% automation rate means, in practice: only about every sixth email lands on the team’s desk, the rest is resolved fully automatically. That creates stability during load peaks, because myphotobook does not need to hire immediately when seasonal volume rises. The freed-up capacity flows to where empathy and experience count: complex complaints and emotional customer cases. For the metrics worth tracking in your own evaluation, see our guide to the most important chatbot KPIs.

Security: no hallucinations, full data sovereignty

Any executive evaluating AI in support faces two questions sooner or later: Is this legally sound? And what happens if the AI sends something that should never have gone out? With OMQ, both questions are addressed by design.

The AI answers exclusively from the company’s stored knowledge. Misinformation through free generation is structurally excluded, and brand control stays with the organisation. At the same time, the solution is GDPR-compliant, the baseline requirement for any digital business model in the European market. The knowledge setup itself is deliberately low-threshold: websites are synchronised via URL, documents simply uploaded. There is no complicated AI training, so teams without IT expertise can start, and the evaluation holds up in front of legal and compliance stakeholders.

5 steps to customer service email automation

What can other organisations take from the myphotobook story? The path condenses into five steps:

1. Identify recurring enquiries: Analyse your inbox and ticket system. Order status, change requests and format questions are typical high-volume candidates. Our article on FAQ automation shows how to approach this systematically.

2. Build a central knowledge base: Maintain service knowledge once, centrally, instead of in silos per channel. With website synchronisation and document import, the setup at OMQ is largely automatic.

3. Start in draft mode: Let the AI create answer suggestions your team reviews first. Confidence in answer quality grows without loss of control.

4. Extend auto-reply step by step: Switch topics with consistently high answer quality to full automation and track your ticket deflection rate.

5. Expand to further channels: The same knowledge base also powers chatbot, help page and contact form. Our guide on how to create a chatbot shows how to get started there.

Conclusion

The myphotobook story shows that customer service email automation is no longer a future topic but proven practice with measurable results: an 83% automation rate, around €408,000 saved per year, and a team that can finally focus on the cases where empathy and judgement make the difference. The key is a connected service ecosystem with one central knowledge base instead of isolated point solutions. Nobody has to choose between scale and service quality: with the right automation strategy, you get both. And the story continues: OMQ Reply is already live for two of myphotobook’s brands, with the rollout to additional brands and countries in preparation.

Frequently asked questions (FAQ)