Most men buy exactly one engagement ring in their life, with no prior experience, on a significant budget, under time pressure, in a category built around making you feel underqualified.
That is a bad combination, and it is why so many people either overspend on the wrong thing or freeze and buy whatever the first showroom put in front of them.
The good news is that the actual decision set is small. There are about six things that matter and a great deal of noise around them. Here is the short version.
Start with the budget and ignore the rule
The two months salary guidance was written by an advertising agency during the Depression and repeated until people mistook it for tradition. It has no basis in anything.
Decide what you are comfortable spending, before you look at anything, and tell the jeweller that number early. Not because they need it to sell you something, but because a good one will use it to show you a realistic set rather than walking you up a ladder.
Be aware that walking in without a number is how you end up spending more than you intended. Showrooms are designed around exactly that.
The four Cs correctly ranked
You will be told cut, colour, clarity and carat matter equally. They do not, and this ranking will save you a meaningful amount of money.
Cut matters most by a wide margin. Cut is not the shape. It is how accurately the stone is proportioned and polished, and it decides whether light entering the top comes back out through the top or leaks away through the bottom. A superbly cut stone with mediocre colour will out-sparkle a poorly cut stone with perfect colour, in every light, permanently. It is also the C that gets the least airtime, because it is hard to make sound impressive on a ticket.
Carat is the one you are most likely to overpay for. Prices jump at the round numbers because that is where demand clusters. A 0.90 carat stone costs noticeably less than a 1.00 carat stone and the difference is invisible on a hand. Sitting just under a threshold is the single most effective saving in this whole process.
Colour stops mattering fairly quickly. The scale runs from D, completely colourless, down through the alphabet. Past roughly G or H, nobody detects warmth without a comparison stone under laboratory lighting. Colour also reads differently against different metals, so a slightly warmer stone looks perfectly white set in yellow gold.
Clarity matters least. The only question that counts is whether you can see inclusions without magnification. VS2 and SI1 stones are usually eye-clean and cost far less than a flawless stone that looks identical at arm’s length. Where the inclusion sits matters more than the grade, since something tucked under a claw is irrelevant.
The rule that serves most buyers: prioritise cut, sit just under the carat thresholds, accept G or H, accept eye-clean, and put everything you save into the quality of the cutting.
The decision you actually have to make first
Two things changed in the last eighteen months and most online advice predates them.
Laboratory-grown diamond prices have collapsed. Manufacturing capacity expanded enormously, average prices fell 20 to 30 per cent against 2024 alone, and some categories are down more than 90 per cent from their peak. Above a carat, lab-grown stones now commonly sell at 5 to 10 per cent of a natural equivalent.
And in late 2025 the Gemological Institute of America stopped grading laboratory-grown diamonds on the D to Z colour and flawless to included clarity scales used for natural stones, replacing them with two descriptive categories, premium and standard.
So the first decision is not size or shape. It is natural or laboratory-grown.
Lab-grown gets you dramatically more stone for the money and has essentially no resale value. Natural costs far more and retains some. Both are real diamonds with identical physical properties. Neither choice is wrong.
What is wrong is not knowing which you are being shown, so ask directly and ask to see the report, because after the grading change the certificate format no longer tells you at a glance.
Certification briefly
Any natural stone of significance should come with an independent laboratory report. GIA and AGS are the two names that carry weight internationally.
An in-house certificate issued by the seller is not a third-party document, for obvious reasons.
Check the report number against the laboratory’s online database before you pay. It takes thirty seconds.
Getting the size without asking
The perennial problem. Options in order of reliability.
Borrow a ring she already wears on that finger and take it to a jeweller. By far the most accurate.
Trace the inside of one on paper. Less accurate, workable.
Ask a friend, sister or her mother. They have almost certainly discussed it already and will be delighted.
Buy slightly large. A ring that is too big can be worn for a few days and resized afterwards. A ring that is too small cannot be worn at all, which is conspicuous in every photograph taken that day.
Or propose with a placeholder and commission the real ring together. Increasingly common, removes all guessing, and suits the very large number of women who have strong opinions about what they want to wear every day for fifty years.
Before you commit, ask what the resizing allowance is. Plain and solitaire bands resize easily. Pave-set shoulders resize within a limited range. Full eternity bands generally cannot be resized at all.
Choose the setting for her life not the display case
This is where you can add more value than anywhere else, because you know things about her a catalogue does not.
If she works with her hands, in healthcare, in a trade, with equipment, or if she is an active person generally, look at bezel settings and low profiles. A tall claw setting is a lever that catches on everything and takes impact directly.
Look at what she already wears. Metal colour, scale, whether she wears rings at all. A large ring on someone who does not wear jewellery is a daily irritation presented as a gift.
Solitaire is the safe default and the easiest to match with a wedding band later.
Halo settings ring the centre stone with small diamonds so it reads considerably larger. Excellent value engineering and slightly higher maintenance.
Bezel is the most secure and increasingly the most fashionable, which is a rare overlap.
Micro-pave bands are beautiful and high maintenance, with small stones that occasionally need replacing.
One question worth asking that nobody thinks of: how will a wedding band sit against this? A ring with a raised setting or curved shoulders may need a contoured band made specifically, which is a custom job. A good jeweller designs for it from the start.
Timing
This is where most of the stress comes from and it is entirely avoidable.
Stock rings can be bought today. Anything made to order runs four to eight weeks, longer if a particular stone needs sourcing. Add a fortnight on top for resizing, the valuation appointment and your insurer actually adding the item to your policy.
So a proposal in three months is comfortable. A proposal in five weeks means buying whatever exists.
Studios that make rather than retail handle interstate commissions routinely now, using video calls, stone options sent as photographs and film, and a posted silver or resin prototype so the fit can be checked before anything is cast. Looking through the Stelios Jewellers engagement rings made to commission is a reasonable way to see what going bespoke actually gets you against a chain equivalent, which is usually more ring for the same money rather than less.
Insurance which everybody forgets
Most home and contents policies cap jewellery cover at a modest figure per item unless the piece is listed separately, and an engagement ring almost always exceeds that cap.
To specify it you need a valuation, and a valuation is a different document from the grading report. The report describes the stone. The valuation states replacement cost. Insurers want the second one.
Do this before the ring leaves your possession for a weekend away. A ring collected on Friday for a Saturday proposal has spent that weekend uninsured, which people only think about afterwards.
Shape and what it costs you
Shape is the most visible decision and it has a price consequence people rarely see coming.
Round brilliants carry a premium of roughly 25 to 30 per cent over other shapes at equivalent quality. Two reasons: more of the rough crystal is lost in cutting, and demand is highest. They are also the most forgiving, because the modern round is the most mathematically optimised cut there is.
Ovals look larger than their carat weight because of the elongated outline, and they are the most requested alternative shape right now. Ask about the bow tie, the dark band across the centre caused by light leakage. Almost every oval has one, the question is how visible, and no certificate will tell you. You have to look.
Emerald and Asscher cuts use flat step facets rather than the triangular facets of a brilliant, producing broad flashes rather than rapid sparkle. They are less forgiving of inclusions because you see straight into the stone, so eye-clean matters more here than anywhere else.
Pear and marquise need careful symmetry or they look subtly wrong in a way people notice without being able to explain.
Princess and radiant hide more weight in the depth of the stone, so a princess can look smaller face-up than its carat weight suggests.
If budget is the constraint, choosing any shape other than round is the second most effective saving available after sitting under the carat thresholds.
Where to buy
Three routes, with real trade-offs.
Chain retailers give you convenience, stock you can try on today and finance options. You are paying for the retail network and the unsold stock sitting in the cases, which is a genuine cost recovered in the price.
Online sellers offer the sharpest pricing and enormous selection. You cannot see the stone, which matters for cut quality and for anything with a bow tie, and returns are a process rather than a conversation.
Independent studios make to order, so you pay for the stone you chose and the metal used rather than financing inventory. Frequently better value than assumed. The cost is time.
Questions that sort a jeweller in four minutes
Where is this ring physically made? A workshop answers instantly and specifically. Vagueness is the single most reliable warning sign in the trade.
Do you do repairs, resizing and setting on site? This tells you whether there is a bench behind the counter.
Is this natural or laboratory-grown, and can I see the report?
What is the cut grade?
Can this setting be resized, and by how much?
What is the warranty, and does it cover stone loss? Usually it does not, which is what annual claw checks are for.
Things not to worry about
The proposal does not need to be elaborate. The overwhelming majority of people asked what they remember describe how it felt, not the staging.
You do not need to surprise her with the ring. Plenty of couples choose it together and keep the proposal itself secret, which is the better division of surprise and is now completely normal.
A smaller stone is not a failure. A well cut 0.7 carat stone in a beautiful setting looks better than a poorly cut 1.2 carat stone in a generic one, and everyone who looks at her hand will see the second thing and not the paperwork.
The short version
Set a budget before you shop. Decide natural or lab-grown and ask to see the report. Prioritise cut over size and sit just under the carat thresholds. Choose the setting for her hands and her job. Start three months out. Insure it before you travel with it.
That is genuinely the whole thing. Everything else is detail, and most of the anxiety around this purchase is manufactured by people who benefit from you feeling out of your depth.
This article is general information only. Diamond prices and grading standards change. Confirm current specifications, certification and insurance requirements with your jeweller and insurer.
