A new revision needs sending, an approval is pending, three people are copied into the thread, and nobody can find the latest decision.
By midday there will be many more emails and the one you need will be hard to find.
There are good reasons to use email. But if all correspondence, drawings and outstanding matters are being held ‘on email’, it can become very time consuming to establish what applies.
How can teams reduce their reliance on email? Each email should ideally be sent to achieve a specific outcome and technology can assist with locating that information afterwards.
What you’ll learn:
- Why Email Becomes a Bottleneck
- Keep Project Decisions in One Place
- Let Automation Handle the Filing
- Where AI Helps and Where It Doesn’t
- Give Each Message a Clear Job
- Start Small and Test the Process
- Wrapping It Up
Why Email Becomes a Bottleneck
As mentioned above, it’s very convenient to send a query to others with relevant attachments and to include anyone else who may be able to respond.
Microsoft’s 2025 analysis of Microsoft 365 activity found that the average worker in its dataset received 117 emails a day.
As well as volume, it’s common for the correspondence and resulting information to be held separately.
For example, a design team may email out an updated drawing, which is approved by a member of management. However, a representative on site will continue to refer to the previous attachment. So although all parties have corresponded, those implementing the works may be working off outdated information.
A 2025 study in Information & Management linked email overload at work with higher job stress and lower job satisfaction.
Keep Project Decisions in One Place
So for information that others may need to refer to, where should it be held?
As discussed, relying on an individual’s email may not be appropriate, for example, they may no longer be involved with the project, be absent from work or purge their inbox. Likewise, saving just the attachments may mean the associated information is unavailable.
With suitable software, such as a construction document management system, emails can be filed against the relevant project and easily located. Others with permission can then access this information to establish, for example, the latest iteration of a drawing and its approval status.
However, as noted above, further information may need to be readily available, such as the reason for revisions and who should ensure the information is updated.
When considering how technology could support the construction industry, In September 2026, the Chartered Institute of Building highlighted that data held in disparate locations and formats prevented greater benefit being achieved. It therefore recommended, amongst other things, having nominated individuals responsible for the information.
Let Automation Handle the Filing
It would be straightforward to file appropriate emails as they are received. But after dealing with a number of correspondences it’s easy to overlook.
These would be applicable where there are frequent occurrences and set parameters, such as:
- Direct emails to appropriate lists, for example, based on project or subject matter;
- automatically file certain types of correspondence and attachments with applicable drawings or documents;
- prompt an individual to action anything outstanding.
Care would need to be taken to only apply these to relevant emails. For example, entering an incorrect project number could result in information being associated with another project, and prompting all users would be counter-productive.
Therefore, it would be prudent to trial the functionality with typical scenarios, including any anomalies such as missing information or several attachments, ensuring users identify any issues.
Where AI Helps and Where It Doesn’t
Have you ever tried to establish what, if anything, was agreed following a length of correspondence?
Various functions are available using AI, for example, to provide an overview of the content, draft a response and identify items to action.
However, these should only be used to assist, as they may omit relevant detail. As an example, a response could state a drawing is approved subject to something else, which an AI generated summary may fail to highlight. So users would need to refer to the original emails and associated documents.
As such, any information provided should be verified, including individuals, dates and attached files. Obvious care also needs to be taken with what information is entered into various platforms.
Ideally, AI would direct users to appropriate information, rather than generate responses and assume authority.
Give Each Message a Clear Job
Technology alone will not prevent all unnecessary emails being circulated. Users would need to agree what information is appropriately sent by email.
It would be quicker to raise a query via a chat function; any actions could be logged with a completion date and responsible individual; and correspondences would continue to be issued but appropriately filed.
Of course, it’s beneficial to have rapid discussion via chat functions, but it’s undesirable for relevant matters to be lost within a large volume of messages.
Some general good practice to minimize unnecessary emails:
- Include appropriate reference, typically the project and what is being requested;
- Only address those who can respond or action anything;
- Indicate a date by which you would like a response, as opposed to issuing all as ‘high priority’;
- Respond to confirm completion or decision and ensure this is readily available.
Naturally this doesn’t seek to impose onerous requirements on all correspondence, just aim to minimize repetition and make clear what action, if any, is required.
Start Small and Test the Process
Any additional process or system would need to be practical to use on a day-to-day basis, avoiding the requirement to continually refer to other platforms.
Identify an area that is frequently problematic. This could include approvals, requests for information or other changes. Track several examples through to completion and identify shortcomings.
Implement as little change as possible to alleviate the issue, for example, agreeing a location to file relevant information, responsibility for doing so and what should be filed. Users would then be able to easily follow this using existing functionality.
Following a suitable period, test if others would be able to readily identify applicable information and associated records.
If not, investigate at what point information is unavailable. Alternatively, if successful, repeat for other scenarios.
Wrapping It Up
Receiving a high volume of emails is undesirable in itself. But then having to request information which should already be known is even more annoying.
Using appropriate technology to hold common information, automate processes and make effective use of AI would alleviate some of this. Combined with basic principles on issuing and recording emails, it should reduce the amount of follow-up required.
