Collect
Field collection with travel, online surveys, web and document sources, sensor readings, interviews and observation studies. The sampling frame is agreed in writing, then a share of every collector's work is independently re-contacted.
Data collection, mining and supply — worldwide
You need the data. We go and get it.
Data Tune collects data in the field and online, mines it for patterns, and hands over something a decision can actually be made on. We work with researchers, market research firms, NGOs and businesses in any country.
Collection is the expertise, not a side service. We design the instrument, define the sampling, travel to where the data is, and document the protocol — so the dataset survives a reviewer, an auditor or a sceptical analyst.
In one paragraph
Data Tune is a data collection company. We gather data that does not yet exist in a usable form — by field survey, interview, observation, sensor, web and document sources — then clean it, structure it, mine it and deliver it with the documentation that lets somebody else judge it. We also build the pipelines and dashboards that keep it arriving, and write the insight reports that say what it means.
What we do
Four disciplines, commissioned together or alone
Most projects touch more than one. The difference between a dataset you can use and one you cannot is almost always the method — so each card says what the work actually is, not what it is called.
Mine
Classification, clustering, prediction, text and web mining. Market and competitor intelligence, price movement, sentiment from review aggregation. Cleaning, normalisation, entity resolution and enrichment.
Build
Ingestion pipelines that handle authentication, rate limits, schema changes and error alerting. Warehouse design and scheduled refresh. Sensor hardware where the data does not exist until you measure it.
Visualise
Dashboards built around the questions your team actually needs answered, pulling from live databases, APIs and sensor streams. Periodic reporting generated programmatically, so a weekly cadence is sustainable.
Method
Six things settled before anything is priced
Choosing the wrong collection method is the most expensive mistake in a data project, and it is almost always made before any money is spent. None of it is visible in the delivered file — so it is settled first, in writing.
1
The question
What you will do differently once you know the answer. This sets the accuracy the data has to reach, and therefore the method and the cost.
2
The population and the frame
Who the data represents and how we reach them. A frame that quietly excludes part of the population is the commonest source of a wrong conclusion.
3
Sample size and confidence
Calculated against the decision, not picked to fit a budget. We say so when the sample you can afford will not support the claim you want to make.
4
The instrument
Questionnaire, observation schedule or extraction specification, reviewed for leading questions, ambiguity and translation loss before it reaches anyone.
5
Validation and back-checks
A share of completed work independently re-contacted and verified. Without this, a collection project is unauditable.
6
Documentation
Protocol, timestamps, provenance, substitutions, response rates and limitations — prepared as the work proceeds rather than reconstructed afterwards.
We will say so
If your question cannot be answered to the confidence your decision needs, at any budget, you hear that during scoping — not in a report with a precise-looking number you should not act on.
Field capability
We travel to where the data is
A great deal of useful data is not online and never will be. Prices on a shelf. Conditions in a field. How people behave in a queue. What a machine is really doing. What someone will tell a trained interviewer but not a web form. For that, somebody has to go.
Our collection teams travel to the site, the region or the country, working to the agreed protocol — household and consumer surveys, retail audits and shelf checks, infrastructure inspections, environmental and agricultural readings, intercept interviews, focus groups, observation studies and sensor placement, with photographic and GPS verification where location matters.
Where a project spans countries we are open about who does the work. Collection by our own teams is identified as such, and where a local partner agency is engaged that partner is named before you commit, with their cost shown as its own line rather than folded into a margin.
Clients
Who we work with
Researchers and academic teams
Instrument design, sampling and collection at publication-grade rigour, documented for peer review. Agriculture, public health, linguistics, economics, environmental science. Fieldwork starts only once written ethics approval exists, and we do not write anyone's thesis.
Market research firms
Capacity when your own team is full, or reach into regions you do not cover. White-label under your branding, to your instrument and your quality standards, delivered in the format your analysts already use.
Businesses and operators
Competitor and price intelligence, market entry research, operational data and the dashboards that make it usable. Usually continuous rather than one-off, because a snapshot answers this month's question and not next month's.
AI and machine learning teams
Training data built to your specification, or drawn from what we already hold. Labelled, balanced, split into train and test, with class-balance reporting and provenance documentation you can defend later.
Continuous supply
Most of the value arrives when the data keeps arriving
A snapshot answers one question once. A supply answers the same question every week and lets you see what moved. We run continuous collection to a schedule, delivered wherever you need it — a cloud bucket, your own warehouse, an authenticated endpoint, a shared sheet, a dashboard. Comparability across waves is designed in from the start.
The part clients underestimate is maintenance. Over a year a website redesigns, an export shifts a column, an API deprecates a field, a source starts blocking automated access. A continuous arrangement covers somebody noticing and fixing that before you find out from a broken report.
Where it can, the work runs inside your own cloud account under your credentials. You are never locked in: you leave with the data, the configuration and the documentation.
Governance
How we work lawfully, and the work we decline
All data is gathered through lawful, transparent and consent-based methods, with the collection scope agreed before work begins. Our practice is aligned to the frameworks that matter to our clients and their data subjects.
| Jurisdiction | Framework |
|---|---|
| European Union | General Data Protection Regulation |
| Australia | Privacy Act 1988 and the Australian Privacy Principles |
| New Zealand | Privacy Act 2020 and the Information Privacy Principles |
| Singapore | Personal Data Protection Act |
| Thailand | Personal Data Protection Act |
| United States | State frameworks, including the California Consumer Privacy Act |
| Sri Lanka | Personal Data Protection Act, No. 9 of 2022 |
In practice: data minimisation, secure handling, anonymisation of personal identifiers where applicable, defined retention, and confidentiality by default. We sign your own processor agreement rather than press a template on you. This describes alignment with recognised frameworks, not a certification or a legal opinion.
Work we decline
Published here rather than discovered mid-project. Each is declined regardless of the budget attached, because the liability lands on you.
- Circumventing logins, paywalls or access controls.
- Scraping personal data from social platforms for marketing.
- Consumer contact lists for unsolicited approach.
- Fieldwork on an academic study before written ethics approval exists.
- Writing anyone's dissertation.
- Conclusions the data does not support, including when they would be the more welcome finding.
Accountability
Who is answerable for the work
Before you hire anyone
Questions buyers ask first
Which countries do you work in?
Anywhere. Online collection, mining, dataset construction, pipelines, dashboards and reports are delivered remotely and are not limited by geography. For field collection we travel to where the data is, and we are explicit about staffing: our own teams are identified as such, and a local partner agency is named before you commit with its cost shown separately.
What makes a data collection company worth paying for?
The method, almost entirely. Anyone can return rows; the question is whether those rows represent what you think they represent. That depends on the sampling frame, the instrument, the validation, and whether anyone back-checked the work — none of which is visible in the delivered file. A supplier who cannot produce that documentation leaves you no way to tell good data from plausible-looking data.
Can you collect data on any subject?
Within the limits published above, yes — the method transfers across subjects more than people expect. We have worked across agriculture, public health, linguistics, economics, environmental science, retail and telecommunications, and the discipline is the same each time. Where a project needs genuine subject expertise we say so and work alongside your specialists rather than pretend to it.
Do you only do one-off projects, or ongoing supply?
Both, and ongoing supply is where most of the value is. A one-off dataset answers a question once; a continuous arrangement answers the same question every week and shows what moved. Arrangements cancel with notice and you leave with the data, the configuration and the documentation.
How do I get a price?
Describe the project and we prepare a quotation against a written scope. Every project is quoted individually because the method drives the cost — the same number of records can be cheap or expensive depending on how it has to be collected and how accurate it must be. The scoping conversation is free and carries no obligation, and indicative rates for standard work are on the pricing page.