← All notes

Notes · · 2 min read

A spec sheet is a data model

HırdavatPro compares professional power tools. The comparison table is not the hard part; the normalisation underneath it is.

A buyer's question sounds simple: which of these two drills should I get? The manufacturer's answer arrives as a PDF, a marketplace listing and a product page, each with different units, different names for the same measurement and a different idea of which fields are worth publishing. Before anything can be compared, it has to be made comparable.

Normalise before you compare

Every product in HırdavatPro maps to the same specification fields. That is what makes a fair side-by-side comparison possible at all. The work is not glamorous: choose the fields, define the unit, decide what counts as the same thing across brands, and keep the mapping when a new model arrives.

Missing is not zero

A specification that was not published is not the same as a specification that is zero. A model that treats a blank torque figure as 0 Nm will rank products by what the manufacturer omitted, not by what the tool can do. The data model needs an explicit “not published” state, and the table needs to show it.

Side by side beats a score

HırdavatPro puts the specifications side by side instead of collapsing them into a single invented score. A score hides the trade-off the buyer is actually making: more torque or less weight, a bigger battery platform or a lower price. The table lets the reader see the trade-off and decide which side matters for their work.

Guides carry the decision

A comparison table answers “what is different?”. A technical guide answers “which difference matters for my job?”. The two belong together: the table is the evidence, the guide is the interpretation. Referral links and lead capture pay for the work, but they are not the argument.

The archive is alive

Tools are discontinued, replaced and re-released under a new model number. Specs change, prices move and stock disappears. A comparison site is not a one-off document; it is a dataset that has to survive the next product cycle. One field change in the model should update every table that uses it.

The table is the easy part to look at. The model underneath is what makes it true.