Contents
- Four channels, one team, one full inbox
- Fast replies bring more back
- The method: one working day, two numbers
- Three findings, and only one argues for a hire
- Worked example: from emails per day to hours per month
- What an agent prepares, and where approval stays
- Unofficial AI use belongs in the same count
- The first step: count one day, then decide
- Frequently asked questions
Key points
- For one ordinary working day, count two numbers: the emails that come in and the replies that go out the same day.
- The two numbers, plus a reason for every email still open in the evening, give one of three findings: volume, a queue before approval, or boomerangs.
- Only with the finding Volume is another hire the obvious answer. With a queue or boomerangs, another person writes replies that wait or come back as well.
- The minutes per email decide the monthly hours. In the worked example with 60 emails a day, one extra minute per email adds 20 hours a month.
- An agent can assign requests, attach documents and draft the reply and the follow-up task. Whatever goes out is approved by a person on your team.
Four channels, one team, one full inbox
Requests from tenants and owners arrive by email, by phone, through a portal and still by letter. An article in Immobilia, the magazine of the Swiss real estate association SVIT (in German), puts the share of tenant requests that still come in by phone at an estimated 40%, despite apps, portals and online damage report forms. The same article describes owners expecting tenant requests to be answered within 48 hours.
When the inbox gets too full, one answer seems obvious: another hire. Sometimes it is the right one. The HSLU real estate blog (in German), however, locates the problem in property management less in the amount of communication than in the lack of steering. Whether that applies to your firm can be counted in one day.
Fast replies bring more back
“When he replies quickly, he immediately has 15 more emails in his inbox.”
From the senior management of a property management firm, translated from German
The sentence describes a pattern many property managers know. A fast reply that does not fully settle the question comes back as the next email: a request for documents, a follow-up question, an appointment that does not suit. Replying faster then creates more emails, and the inbox looks fuller than before.
For the hiring question this matters. If you do not count boomerangs separately, they sit in the same number as new requests, and faster replies look like more volume. Another person then writes additional replies that come back as well.
The method: one working day, two numbers
You need a tally sheet, a clock and a day on which the inbox runs as usual, meaning no return from holidays and no statement deadline. The person who runs the inbox can count, or you can.
- 01Count emailsOne mark for every email that comes in, split into tenant, owner and other, such as tradespeople or insurers. If it answers a reply sent the same day, it also gets a mark under boomerang.
- 02Count repliesOne mark for every reply that goes out the same day.
- 03Give open emails a reasonWhatever is still unanswered in the evening gets a reason: waiting for approval or waiting for documents, such as the building regulations or the client file.
- 04Time five emailsFor five emails, measure the time from opening to the sent reply and take the average.
- 05Scale up and readEmails per day times minutes per email, divided by 60, times working days in the month. Then divide the same-day replies by the emails.
Count portal messages like emails. Calls and letters can go in a separate column. For the first finding, emails are enough, because that is where replies, waiting times and boomerangs are on record.
Three findings, and only one argues for a hire
| Finding | How you recognise it | What helps next |
|---|---|---|
| Volume | Almost everything goes out the same day. The hours sit in the emails themselves. | Spot recurring replies and prepare them. Here, another person can help. |
| Queue before approval | Many emails wait overnight, with the reason “waiting for approval” or “waiting for documents”. | Clarify who approves and where the documents are. Another person writes replies that wait as well. |
| Boomerang | Fast replies bring new emails on the same matter. | Make replies more complete: attach the documents, name the next step and the date. |
Volume leads to a conversation about capacity. A queue leads to the question of who approves, and with which documents. Many boomerangs put the replies themselves on the table. Before counting you have an impression. After one day you have two numbers.
Worked example: from emails per day to hours per month
A worked example with round assumptions, not a measurement from any firm: 60 emails per working day, 5 minutes per email, 20 working days a month. That makes 300 minutes, so 5 hours a day and 100 hours a month. At 6 minutes per email it is 120 hours. In this example, a single minute per email adds 20 hours a month. That is why you time real emails rather than estimate.
In the same worked example, 39 replies go out by the evening, 65% of the emails received. Of the 21 still open, 14 are waiting for approval and 7 for documents. That is the finding Queue. Enough is being written; the replies are waiting for someone to approve them. In this example, another hire would mostly lengthen the queue in front of the same approval.
What an agent prepares, and where approval stays
An agent for the tenant inbox works on all three findings, each at a different point. Where the hours sit in volume, it drafts the recurring replies. In a queue, the documents people search for today arrive already attached. Against boomerangs, it makes sure the reply settles the question in full, with the attachment and the next step.
The agent prepares
- Assigning each request: property, unit, person, matter, urgency
- Attaching documents: lease, house rules or building regulations, earlier reports
- Drafting the reply in your firm's tone
- Drafting the follow-up task, with owner and deadline
Stays with your team
- Reading, adjusting and sending: no reply goes out without approval
- Anything with legal effect, such as notices of termination, defect notices or commitments on costs
- Instructions to tradespeople above the authority limit
- Borderline cases the rules cannot assign clearly
The rule of thumb from the HSLU real estate blog draws the same line: “high impact = a person must decide” (our translation). Among the areas that call for caution, the article lists external communication with legal effect and automated decisions without control. The SVIT article also describes processes that close simple standard requests without any human intervention. We draw the line more tightly: every reply to a tenant goes through approval, and the log records what the agent proposed and what the person sent.
Unofficial AI use belongs in the same count
A bachelor's thesis at ZHAW surveyed 72 managers and employees of small property management firms. According to the press release (in German), a third of them use AI even though their firm does not officially use any. In the survey, AI is used mainly for written communication and property listings, and binding rules exist in very few of the firms.
For your count, that means: on counting day, also ask what replies are drafted with today. If tenant data ends up in a private account, that is a data protection question long before it is a question of efficiency. What a simple internal rule should contain is covered in AI policy for SMEs.
The first step: count one day, then decide
Pick an ordinary working day this week and the inbox that tenants and owners write to. Print the free Inbox Hour Counter: tally sheet, scaling and the three findings on a few pages. By the evening you have two numbers and a reason for every open email.
Some replies to owners are work that may go beyond the base fee. Whether it can be billed is decided by the contract, as described in Flagged is not billable. How Grow Amplified first costs a process like this in your own hours, and only then builds it, is described under Approach.
Frequently asked questions
Is one day enough for a reliable number?
For the first finding, yes, as long as it is an ordinary day. If you want to be sure, count a second day in a different week and compare. If the findings differ a lot, a full week is worth it.
How do I count phone calls, portal messages and letters?
Portal messages like emails. Calls and letters in a separate column, with the same question: dealt with the same day or not, and if not, why. For the first finding, emails are enough.
What is a boomerang?
An email that follows a reply sent the same day, because the reply did not fully settle the question: a follow-up question, a missing document, an appointment that does not suit. Count boomerangs separately, otherwise faster replies look like more volume.
May an AI agent answer tenant requests on its own?
Technically, yes. We recommend approval for every reply, because tenant communication often has legal effect and tenant data is personal data. The Federal Data Protection and Information Commissioner (in German) states that the Data Protection Act applies directly to AI and that people have a right to know whether they are corresponding with a machine.
Does an agent replace the extra hire?
That depends on the finding. With a queue or boomerangs, an agent works where another person changes little. With the finding Volume, a hire can be the right answer, and an agent takes the sorting and searching off that person's desk.
15 years in banking and consulting, in transformation and finance operations. Today he builds review lanes with Grow Amplified: agents read and reconcile, a person approves, and a replay log records who decided what.