Pipeline Flowgrid
Flagship study

Radio-access performance study

A three-to-six-week reading of one market or technology layer — busy-hour maps, a ranked worst-cell list, and a written note for the capacity meeting.

From 12,400 GEL

Request this study

Paper map with marked routes on a wooden table

A radio-access performance study is the work most planning desks actually need and rarely have time to finish in-house. Counters exist. The meeting still opens with a nationwide average that conceals the Batumi boulevard on a Friday in August and a ridge site that has been quietly failing since March.

We take one market or one technology layer and stay there. The pack is built for a named sitting — a capacity freeze, a refarm rehearsal, a conversation with finance about sector adds — not for a shelf of unread dashboards.

The first week is unglamorous on purpose. Georgian networks accumulate renamed sites, microwave hops that still carry a 2G identity in a spreadsheet, and outage hours that were never stripped from a busy-hour export. If we skip that reconciliation, the congestion map will accuse the wrong rooftop. Levan’s planning group in western Georgia learned this the irritating way: we spent days on the list before a single choropleth appeared. That is the job.

Maps in the pack are clustered at the hours you name. A tourist market in Adjara is not described by a Tuesday at 11:00. We will ask for Friday evening, Saturday afternoon, and a quiet midweek window so the seasonal load is visible next to the ordinary day. Worst-cell lists carry the counters that earned each row — PRB utilisation, retainability, setup failure, or whatever the vendor file actually contains — rather than a composite score nobody can audit in the room.

The written note is short. Each theme gets a finding, the cells that illustrate it, and a sentence on what would change the picture (a sector add, a backhaul lift, a parameter you already know is frozen until a swap). We do not issue a work order. Your radio engineers remain the people who touch the live layer.

Pricing starts at 12,400 GEL for a single-city 4G cut with a usable site list. A regional layer, a dual-technology reading, or a pack that must also reconstruct a named incident is quoted after we see the file list. The briefing is part of the study fee when it is held in Batumi; sittings elsewhere in Georgia add travel at cost, stated in the estimate.

If you already know the meeting date, put it on the request. A pack that arrives the morning of the freeze is a different (and usually worse) piece of work than a pack the room has had three days to mark.

Who it is for

RAN planning leads, cluster owners, and operations managers at mobile operators or fixed-wireless providers who need one market or one technology layer explained before a capacity freeze, a spectrum refarm, or a board note.

What you leave with

A bound written study (typically 10–18 pages), cluster maps at the hours you name, a ranked cell list with the counters that put each row there, and one briefing of up to half a day.

Scope

One market or one coherent cluster (for example Adjara 4G, Batumi metro, or a named highway corridor) and one radio technology layer unless we agree a dual-layer cut in the estimate.

Duration and place

Three to six weeks from a usable data drop, depending on site count and how clean the site list is. Analysis is done in Batumi. The sitting can be at Gorgasali Street, at your operations floor in Georgia, or by video if the planning lead is not in Adjara.

Included

Data intake and a one-page quality note on the files you sent; congestion and coverage-themed maps; worst-cell ranking; written findings; one revision of the pack after the briefing; travel within Batumi for the sitting.

Not included

Live parameter changes, vendor OSS licences, 24-hour monitoring, drive testing we did not quote, and work on a second market without a new estimate.

Preparation

A current site list with CGI or equivalent, counter exports for the agreed window, a known-outage calendar, and a sentence naming the meeting the pack has to serve.

Constraints

We do not take write access to the live radio layer. If the site list and the counters disagree, we stop and send a query rather than inventing identities for cells.

How the weeks run

  1. Intake call and a written list of files we need. If the site list is older than a year, we say so immediately.
  2. Quiet reading week: we reconcile names, drop known outage hours, and send a one-page quality note before any map is drawn.
  3. First maps and a draft worst-cell list. You mark what is already in a swap programme so we do not lecture you about it.
  4. Written findings. One review call, then the pack is frozen for the briefing.
  5. Half-day sitting. One revision of the pack is included if the room catches an identity error or a missing corridor.

Next step

Write with the market, the layer (for example 4G Adjara), the meeting date if you have one, and whether a site list already exists. We reply with a scoped estimate, not a catalogue price, when the cluster is larger than a single city.

Write to the desk