Our methodology: how we guarantee the reliability of every published price

Our methodology: how we guarantee the reliability of every published price

Price data is only valuable if it is reliable. That is the founding principle of Zando. Since our first day of collection at the Total Bacongo market in Brazzaville, we have built a rigorous protocol to ensure that every price published on our platform reflects field reality.

Here, in full transparency, is how we work.

01A trained field agent network

Zando relies on a network of locally recruited collection agents present at markets in Brazzaville and Pointe-Noire. These agents know the markets they cover — some have been working or trading there for years.

Each agent is trained to our collection protocol:

  • Identify the effective transaction price, not the displayed price — the price actually charged during a transaction
  • Use standardized measurement units (kilogram, liter, piece depending on the product)
  • Systematically report: product, market, price, unit, collection date and time
  • Flag any observed anomaly: unusually low or high price, stock shortage, large merchandise arrival

02Double verification

This is the core of our quality protocol. Every price collected by one agent is independently verified by a second agent present at the same market on the same day, without any coordination.

Validation rule:

  • Gap < 10%: the price is validated and published
  • Gap between 10% and 25%: a third agent is mobilized for arbitration
  • Gap > 25%: both readings are rejected and collection is rescheduled

03Automated processing

Data submitted by agents passes through an automated processing pipeline that:

  • Checks submission format consistency
  • Compares the submitted price against recent history for the same product on the same market
  • Detects statistical outliers (prices deviating more than 2 standard deviations from the median)
  • Assigns a reliability score to each observation before publication

04Human moderation

Observations flagged as anomalies by the automatic system are manually reviewed by our team before any publication. Data not validated by this process is never published on the platform.

Update frequency

Field collection

2 to 3 times per week on each covered market

Price publication

Every week, on Monday morning

Monthly price index

On the 5th of each month for the previous month

Price alerts

In real time upon observation validation

What we do not do

We do not publish prices based on estimates or declared prices without an observed transaction

We do not weight prices by trading volumes — our prices are simple averages of validated observations

We do not extrapolate prices to uncovered markets

Our commitments

Transparency

Every published price indicates its collection date, source market, and the number of observations it is based on.

Traceability

Every observation is archived and can be audited.

Continuous improvement

Our protocol evolves quarterly based on field feedback and quality analyses.

Have questions about our methodology? Contact our team — we respond to all transparency requests from institutions and partners.

Contact us