Choose CRM dropdown options that staff can use without guessing
Prepared with AI assistance by Alactic Editorial. Examples identified as hypothetical are illustrative, not customer results.
If staff repeatedly choose the nearest answer in a CRM dropdown, start by reviewing the choices—not by reminding people to complete the field. A usable list asks one clear question, offers distinguishable answers, and gives people an honest route when the information is missing or the case does not fit.
Design the field around the decision its value supports. Then ask different staff members to classify the same cases independently. Their disagreements will expose overlapping labels, missing definitions, and questions that should not be squeezed into a single dropdown.
The filled hypothetical worksheet below shows that process for a customer-request field. It is a fictional worked example, not an Alactic customer result. Its categories, cases, counts, and review arrangements are illustrative; adapt them to the work your team actually performs.
1. Start with the decision, not the existing list
Before editing any options, finish this sentence: “When someone selects this value, we use it to decide…” A useful answer names an action: which team should review a request, which information staff should collect next, or which work queue should receive it. “We might want it later” is not enough to justify a detailed required field.
Also identify the thing being classified. Is it the customer, the current request, or an individual task within that request? A customer can have several unrelated requests. A request can contain several tasks. If users classify different units, even excellent labels will produce inconsistent records.
For the hypothetical worksheet, the field classifies the current customer request. Its purpose is to help staff find the right handling instructions and suggest an initial reviewer. It does not classify the customer's overall relationship with the business.
Write down when staff must answer. A category that becomes obvious after investigation may be impossible to choose at first contact. Either collect it later or allow an explicit unresolved value. Requiring an answer before the evidence exists creates false certainty.
2. Filled review worksheet: diagnose the current choices
Hypothetical worksheet, part A — current design
- Field: Request type.
- Record being classified: One current customer request.
- Decision supported: Find relevant handling instructions and suggest the first reviewer.
- Who selects it: The staff member recording the request.
- When selected: After reading the message or completing the initial conversation.
- Current choices: General enquiry; Account issue; Update details; Billing; Invoice query; Urgent; Other.
- Current gap: No definition of what to do when a request contains unrelated tasks or insufficient information.
The list combines several different questions. “Urgent” describes time sensitivity, not the subject of the request. “Billing” and “Invoice query” overlap. “Account issue” could mean a login problem, an address correction, or almost anything associated with an account. “General enquiry” and “Other” provide two escape routes without explaining their difference.
Do not simply rename these choices and keep the structure. First decide which distinctions change the work. If invoice copies and charge explanations follow the same handling instructions, separate options may add effort without improving the next decision. If they require materially different handling, a split may be useful.
Low usage is a reason to inspect an option, not automatically remove it. A rare case may still need a distinct route. Conversely, a frequently selected option may be popular only because its wording is broad enough to absorb everything.
3. Use disagreements to locate the actual ambiguity
Give reviewers the same case descriptions and ask them to choose independently before discussing the answers. Include straightforward cases, boundary cases, missing information, and requests containing multiple tasks. Do not tell reviewers which answer the field owner expects.
Hypothetical worksheet, part B — initial classification disagreements
- Case A: “Please send a copy of last month's invoice.” Reviewer one selects Invoice query; reviewer two selects Billing. Problem: Both labels plausibly describe the same request.
- Case B: “Our office has moved. Please change our contact address.” Reviewer one selects Update details; reviewer two selects Account issue. Problem: The broader option swallows the more specific one.
- Case C: “I cannot sign in, and I need access before a meeting.” Reviewer one selects Account issue; reviewer two selects Urgent. Problem: The list mixes subject and time sensitivity.
- Case D: “Can someone help with my account?” Reviewer one selects Account issue; reviewer two selects General enquiry. Problem: The message does not contain enough information, but neither option records that fact.
- Case E: “Please update our contact address and explain this extra charge.” Reviewer one selects Update details; reviewer two selects Billing. Problem: The request contains two distinct tasks.
For each disagreement, ask what evidence drove the selection. A reviewer may have inferred facts that were not in the case, followed a local team convention, or interpreted a label differently. Those causes need different fixes.
Training can explain a stable boundary, but it cannot repair a missing category or make two overlapping definitions exclusive. If the explanation requires a long chain of exceptions, simplify the field or reconsider what it is trying to classify.
4. Filled review worksheet: replace labels with usable boundaries
Hypothetical worksheet, part C — revised design
Revised field label: What does this request need help with?
Staff guidance: Choose the subject supported by the customer's message. Do not infer an underlying cause. Record time sensitivity separately. If the request contains unrelated tasks, use the agreed task-separation process rather than silently choosing one.
- Access help: The customer cannot sign in or use an existing account. Includes password and sign-in difficulties. Excludes changing contact information.
- Contact details change: The customer asks to change stored contact information, such as an address or telephone number. Excludes requests to explain a charge.
- Invoice or charge help: The customer needs an invoice document or an explanation of an invoice or charge. These are combined because, in this hypothetical workflow, they use the same initial handling route.
- Information request: The customer asks how an existing service or process works, without reporting an access problem, requesting a contact change, or asking about an invoice or charge.
- Not enough information yet: The subject cannot be determined from the available message. The next action is clarification.
- No suitable category: The subject is understood but does not fit any defined choice. The next action is category review.
Choices removed: Account issue is replaced by narrower, observable subjects. Billing and Invoice query are combined. Urgent is removed from this field. General enquiry is replaced by a bounded information category. Other is split into two different unresolved states.
The distinction between the last two choices matters. “Not enough information yet” means the staff member needs facts. “No suitable category” means the list needs attention. Combining them hides whether the remedy is a customer follow-up or a design change.
Keep essential definitions available where staff work. If the workspace supports field help, use it; otherwise provide a short shared reference. Check the current workspace and plan before relying on any particular configuration feature.
5. Give exceptions an owner and a next action
An exception option is only useful if someone knows what happens next. Without ownership, it becomes a permanent holding area. Without permission to use it honestly, staff return to choosing an inaccurate category.
Hypothetical worksheet, part D — exception route
- Not enough information yet: The request owner asks a focused clarification question and keeps responsibility for the record. Once the subject is known, the owner replaces the unresolved value.
- No suitable category: The request owner records a short description of the known need and sends it to the designated workflow coordinator. The coordinator assigns a handling route; the field owner considers whether the category list needs revision.
- Multiple unrelated tasks: Keep the original message available and separate the work into linked tasks or records if the workspace supports that approach. Otherwise use a documented manual handoff that identifies each task and its owner. Do not discard the secondary request.
- Review responsibility: In this hypothetical design, the field owner reviews exception cases weekly and records whether each needs a new category, a clearer boundary, or no taxonomy change.
Do not create a new option for every exception. A new choice is justified when it describes a repeatable kind of work with a meaningful distinction. A unique request can remain an exception with an explicit owner and handling decision.
There is also a trade-off in splitting work. Separate tasks can make ownership clearer, but duplicate records can confuse staff if links and responsibilities are unclear. Use the simplest arrangement that preserves every requested action.
6. Retest consistency without mistaking agreement for accuracy
Test the revised list using both the original cases and fresh ones. Reusing only the discussed cases can measure memory rather than clarity. Ask reviewers to record a choice and a short reason, then compare results.
Hypothetical worksheet, part E — revised classifications
- Case A, invoice copy: Both reviewers select Invoice or charge help.
- Case B, contact address change: Both select Contact details change.
- Case C, sign-in difficulty before a meeting: Both select Access help; time sensitivity is handled separately.
- Case D, unspecified account help: Both select Not enough information yet and propose a clarification question.
- Case E, address change plus extra charge: Both identify two tasks and use the exception procedure rather than assigning one misleading category.
Track exact agreement on cases that should receive a single category: cases where reviewers selected the same value divided by cases independently classified. Track correct use of the multi-task route separately. Do not count every use of an exception route as a classification failure.
Agreement alone is insufficient. Reviewers can consistently choose a broad but unhelpful answer. Check whether their reasons follow the definitions and whether the resulting handling route makes sense.
For example, a hypothetical team might agree on eight of ten single-category cases. That is a diagnostic count, not a benchmark. The two disagreements matter more than the headline: do they expose one unclear boundary, or do they reveal that the field asks a question staff cannot answer?
7. Change the live field without rewriting history blindly
Before replacing old choices, inspect where the values are used. They may affect saved views, required-field rules, routing, exports, integrations, or staff instructions. This is a dependency check, not an assumption that any particular CRM provides these features.
Make a mapping for old values, but do not force every historical record into a new category. An old “Account issue” value does not reveal whether the request concerned access or contact details. Retain its historical meaning where possible, or mark it as a legacy value rather than inventing specificity.
If the workspace allows retiring a choice from new entry while preserving existing records, assess that approach. Otherwise plan the migration with whoever manages the system. Keep a record of the old labels, new definitions, and change date.
Test any dependent workflow before broad release. If staff discover that a new selection sends work to the wrong place, provide a manual correction route and pause the affected routing change. Fixing category wording should not leave live requests without an owner.
8. Run a small review tomorrow
Choose one troublesome dropdown rather than redesigning every field at once. A practical first review can use this hypothetical test plan:
- Select ten de-identified cases, including unclear and multi-task requests.
- Ask two staff members to classify them independently using the current list.
- Record each disagreement, the evidence used, and the action each choice would trigger.
- Draft revised choices with inclusion boundaries, exclusions, and unresolved routes.
- Retest with fresh cases and check both classification consistency and the proposed next action.
After release, review unresolved values, corrected classifications, requests sent to the wrong team, and questions staff still ask about the field. Record counts alongside their causes. A rise in explicit uncertainty may initially mean people have stopped hiding missing information behind a plausible answer.
The immediate deliverable is a completed worksheet: one question, a bounded set of answers, a named exception owner, and evidence of where users still disagree. That gives you a specific next revision—not another reminder to staff to choose more carefully.
Put the next conversation to work.
Explore Alactic Communications for WhatsApp conversations, campaigns and AI-assisted workflows. Choose the product and plan that fit your team.
Explore Communications