Data collection tools should work smart

1792957313009

Nobody builds a QA team for fun. Tools like KoboToolbox, ODK and SurveyCTO just put data collection on phones, and they did that job well. What they left open was confidence.

During a data collection project every night, someone on the team is tasked with validating the day’s submissions and checks a sample. They play a few audio clips, look at the GPS pins and flag what looks wrong. The rest goes through on trust.

Overtime market research agencies filled the gap with people. Back-checkers, call-backs, review spreadsheets, and a supervisor who knows which enumerator to watch. That is real expertise, and we respect it.

But a spot check proves only that the interviews you looked at were done properly. Say you check one response in ten. Nine go to the client on trust.

Timing is the second problem. The check happens after the enumerator has left the community. If a pin is wrong on Tuesday and you see it on Thursday, a return trip costs fuel and a day of the team’s time. Sometimes the household cannot be found again.

Most enumerators are honest and work hard. But fieldwork carries pressure. The road is bad. The respondent is not at home. The network drops in the middle of an interview. The supervisor wants numbers by evening.

Under that pressure, a form is sometimes completed from the roadside or from the next compound. That is usually a shortcut and not a scheme. It still puts a wrong location into your dataset, and your sample may never touch it.

A good QA team shrinks this risk. It cannot remove it. A sample is still a sample.

The cost that is not on the budget line

Free tools are free to download. They are not free to run. Someone writes the validation rules. Someone designs the back-check sheet. Someone matches the GPS file against the assigned LGAs.

That someone is on your payroll. In a busy month, that is the person who should be speaking to the client. There is also key-person risk. When the QA lead who built the system resigns, much of the system leaves with them.

Are the free tools enough? For collection alone, they may be. KoboToolbox and ODK are good tools, and many teams have built careful workflows around them.

Our point is different. Collecting is the easy part. Proving that each response was taken where and how it should be is where the hours go. These tools record the location of an interview. In many setups, checking that location against the assigned area, for every response, is still left to your team.

Field data quality control without the manual work

On WeCollect, location verification runs on every response as it is submitted. A project can be location-locked, so an interview outside the assigned area is flagged while the enumerator is still there.

Supervisors see coverage by area during the day, not in a spreadsheet at night. Forms work offline and sync when the network returns.

The QA team does not disappear. Its work changes. Instead of hunting for problems, it reviews flagged cases, judges answers that look strange and speaks to enumerators. That is a better use of experienced people.

Clients also ask a harder question than they used to: how do you know the interview happened where you say it did? A dataset with location evidence on every row answers that without a long explanation. It also shortens the argument when a client queries a number.

The agencies that run projects with us come for two things: quality they can defend and delivery at speed. Both depend on finding problems while the team is still in the field.

You can start on the free plan and run a real project before you spend anything. Location verification applies to every response on it. When your volume grows, a plan for 10,000 submissions a month is ₦150,000. For comparison, SurveyCTO lists its 10,000-submission Pro plan at $350 a month, billed monthly. Put either figure beside what your QA hours cost in a busy month.

A fair test

Before you decide, put some questions to your own process. What share of this week’s responses had their location checked against the assigned area? How long passes between an interview and its first review? How many hours did QA take last month? What happens to the process if your QA lead is on leave?

If the answers are comfortable, keep what you have. If they are not, run a test, and do not switch anything. Take one live project and run it on WeCollect beside your current system. Compare the flags. What did your QA team catch? What did the platform catch earlier?

If your system wins, you have lost nothing. If it does not, you will know where the hours are going.

Your team has worked hard long enough. Let the tool work smart: set up the test at https://wecollect.tech

Leave a Comment

Your email address will not be published. Required fields are marked *

Save to list

Scroll to Top