Area 14 · Data, AI and consulting

Data and analytics

Imagine a furniture manufacturer in Kayseri that sells through a dealer network. On the first Monday of each month the sales manager pulls turnover out of Logo, downloads the Trendyol settlement report, subtracts returns and commission by hand, converts the lot into euros and emails it to the directors. Finance turns up with a different figure for the same month, because it applied the exchange rate on the invoice date while sales used the rate on the order date. Half an hour of every meeting goes on deciding whose number is right. That ritual burns several working days a month and leaves the company with two competing versions of the truth. Our job is to reduce them to one: tidy master data, all sources feeding a single store, reports that refresh themselves and metrics everybody defines the same way. The whole engagement runs remotely, over secured access and video calls.

3
services covering this area
1 turnover
figure shared by sales, finance and the board
€55
hourly plus VAT, or fixed once scoping ends
TR / EU
hosting for your data in Turkish or EU data centres

Symptoms of neglected data

Nobody ever raises a ticket titled “master data clean-up”. People request a shiny dashboard; digging into the underlying systems then uncovers something from this list. If a couple of these sound familiar, buying a different reporting product will not fix anything.

Every review starts with a dispute

We pin the definitions down in writing. Sales goes by order date and finance by invoice date; one side nets off returns and the other does not; everyone converts at a different day's rate.

One company, three records

We merge them and block new copies. “Yılmaz Mobilya San. ve Tic. Ltd. Şti.”, “YILMAZ MOBİLYA” and a card with no tax number end up as three distinct buyers in any analysis.

Closing the month takes seven days

Loads become automatic. Exporting, pasting and reconciling occupies your most numerate colleague, right in the week the KDV return is due.

Year-on-year comparisons tell you nothing

We separate nominal from real growth. With high inflation, lira turnover rises every year regardless; what matters is units sold, the foreign-currency equivalent and the inflation-adjusted amount.

Goods-in staff retype delivery paperwork

Machine reading takes over. Anything arriving as e-İrsaliye is fine, but paper notes from small suppliers and line items on invoices from Italy or China are still entered manually, row after row.

Nobody knows what a marketplace really leaves you

We work out true profit per channel. Trendyol, Hepsiburada and your own site are ranked by gross sales rather than by the margin remaining once commission, cargo share, service charges and refunds come off.

A dashboard with forty charts is less useful than one page answering three questions. Reporting projects often kick off with “let us see everything” and end with tabs nobody clicks. We go the other way round and start from the decisions management faces this month. Should a dealer's credit limit go up? Should a product be pulled from a marketplace? Should a raw material be ordered early? Each metric is tied to one of those choices, and if it cannot be tied, it stays off the page. Data quality is measured before any chart is drawn, and the error count is shown to you without cosmetics.

Retiring the manual Monday workbook

Scope begins narrow: usually sales or inventory, with only a handful of indicators. A second subject is added once people rely on the first report every day. Work on your systems happens through encrypted remote sessions, and discussions take place over video.

01

A shared glossary

Together with finance and sales we write down how turnover, gross margin, returns, currency conversion and an “active dealer” are calculated. Without that agreement, report number one gets torn apart at its very first outing.

02

Mapping the sources

Your ERP, online shop, marketplace panels, advertising accounts and those side spreadsheets nobody admits to. Each gets a named owner, a connection method and a note of any fields that must appear in the KVKK personal data inventory.

03

Running in parallel

The new automated view and the old hand-made workbook coexist for several weeks. Every gap between them is explained before the directors let go of the manual file.

04

Growing layer by layer

Fresh topics and metrics reuse the same data model and load jobs, which is why each new report comes in noticeably cheaper than the first.

Frequently asked questions

Usually both. Statutory records and reconciliation with your mali müşavir are in lira, yet a business that exports, buys raw materials in foreign currency or prices in euros struggles to read trends without the converted figure. The glossary fixes once and for all which rate applies and on which date, and the warehouse stores every transaction in lira as well as in the currency you choose.

Companies living in Microsoft 365 with Entra ID sign-ins usually land on Power BI, since access follows the accounts people already have and exporting to Excel is trivial. Teams comfortable with SQL who would rather avoid per-seat licences often prefer a self-hosted Metabase. Looker Studio is an option for Google Workspace users. Whichever you pick reads from one warehouse, and you get a side-by-side cost comparison before committing.

It should not. Extraction runs after hours or in modest chunks, ideally via the vendor's service layer or a replicated database, and heavy queries are kept well away from office hours. Keeping reports on a separate warehouse is precisely what protects daytime work: dashboards hit their own copy while whoever is invoicing carries on undisturbed.

Hosting sits under an account registered to your business, with Hetzner, OVHcloud, IONOS, an EU region of Azure or AWS, or on hardware you already run. Under KVKK, Apply acts as data processor, and a contract sets that out. If the server is abroad, we work through the cross-border transfer question together with your legal adviser. Access is role-based, so the board sees everything while a regional sales rep sees only their own territory, and every access is logged.

The biggest variable is how healthy your records are. Reasonably tidy data means a first report arrives fast and most effort goes on settling definitions. Thousands of duplicates turn the clean-up into the main job, so dates are promised only after we have inspected what you hold. Billing is either a fixed price or €55 per hour plus VAT with an agreed cap on hours.

Not at all. Excel is a superb analysis tool; the problem is copying data into it by hand every month. Connected through Power BI or straight to the warehouse, a workbook opens with current figures that match what everyone else sees. Personal pivot tables live on, they simply draw from a common source.

Let us bring order to your figures

Tell us which reports are still assembled by hand and where the numbers fail to agree. We will review your sources and point out where it makes sense to begin.

Availability
Weekdays 09:00-18:00 Turkey time (GMT+3); an answer follows by the next working day
Calls
By video, over Microsoft Teams or Google Meet

The only cookies here are the essential ones: they keep the site running and remember the city you picked. Nothing is used for advertising or tracking. See our privacy notice for more.