What an "All-Time High" Offer Really Means: The CP Offer Score, Explained

Nearly every card-comparison site will tell you an offer is "all-time high" this month. Most of them have no idea what the offer looked like a year ago — they're guessing, or worse, repeating a claim from another site that was also guessing. CreditPoints tracks actual offer history per card, and our own internal rule is stricter than the marketing language you'll see elsewhere: we never say "all-time." We say "highest since tracking began," we date it, and we label whether we personally observed it or reconstructed it from public archives.

This is the guide to how the CP Offer Score actually works — the five factors, the weights, and the difference between "elevated," "near-high," and a genuine tracked high. We'll walk a real card through the full calculation with real numbers from our own tracking data. If you haven't read how welcome bonuses actually work or how to hit minimum spend without overspending, those two guides cover the mechanics this one assumes you understand.

Quick answer

The CP Offer Score is a 0–100 rating built from five weighted factors: how the current offer compares to that specific card's own tracked history (20%), net first-year value after the annual fee (25%), ongoing "keeper" value once the bonus is gone (15%), how hard the card is to get approved for (15%), and how well its earning categories match everyday spending (10%). A card scoring 85+ is "Exceptional"; 70–84 is "Very Strong"; 55–69 is "Solid"; below 40 is weak. We never claim an offer is the best ever — only the best we've observed since we started tracking, with the date and source attached.

• Historical Offer Strength (20%) — is this offer above, at, or below what we've tracked for this exact card?
• First-Year ROI (25%) — after the annual fee, what's the net dollar value in year one?
• Keeper Value (15%) — is the card still worth holding once the bonus disappears?
• Approval Accessibility (15%) — how competitive is approval for a typical applicant?
• Spend Alignment (10%) — do the earning categories match how people actually spend?

💡 Pro tip — When you see "all-time high" on any site, ask: since when? A site that started tracking offers in 2024 cannot honestly claim a 2016 offer wasn't higher. CreditPoints labels every high as either "observed" (we tracked it ourselves) or "archival" (reconstructed from public records) — ask any comparison site whether they do the same.

Key takeaway: the CP Offer Score is a weighted five-factor calculation, not a single "is this a record" flag — and even the historical-strength factor is phrased as "highest since tracking," never "all-time."

Why "all-time high" is nearly always the wrong claim

Most credit-card sites don't maintain their own historical offer database. They see today's number, compare it to a vague memory of "what this card usually offers," and print "record-breaking" because it drives clicks. The claim is unfalsifiable to the reader and frequently wrong — plenty of "all-time high" claims get quietly walked back a few months later when the same card runs an even bigger offer.

CreditPoints' internal rule, built directly into our offer-tracking engine, is deliberately narrower: never claim "all-time" — use "highest since tracking started," dated, with provenance. Every historical high we cite carries one of two labels — observed (we tracked it ourselves in real time, through our scraper and admin verification pipeline) or archival (reconstructed from public records like Wayback Machine snapshots predating our own tracking, flagged explicitly rather than implied as something we personally watched).

That distinction is why our language is more conservative than most competitors' — "highest since tracking" is a verifiable fact; "all-time high" almost never is.

Key takeaway: "all-time high" is usually an unverifiable claim. CreditPoints tracks per-card offer history with a date and a provenance label — "highest since tracking" is a fact; "all-time high" almost never is.

The five factors, in detail

Historical Offer Strength (20%)

We classify every tracked card's current offer into one of five tiers, purely based on that card's own history — never compared to a different card:

This is the factor most people mean when they ask "is this a good time to apply" — and it's deliberately the smallest-weighted factor that still moves the needle, because a big number means nothing if the card charges an annual fee the bonus doesn't come close to covering.

First-Year ROI (25%) and Keeper Value (15%)

ROI is net first-year value (bonus plus realistic credit usage plus estimated category earnings, minus the annual fee) divided by the fee. A card returning 10x its fee in year one scores 100; below breakeven scores 10–25. Keeper Value asks the harder question: strip out the signup bonus — is what's left (credits, ongoing earn rate) still worth the fee every year after? A card can post an Exceptional signup-year ROI and a weak Keeper Value if the bonus is its only real draw.

Approval Accessibility (15%) and Spend Alignment (10%)

A 200,000-point bonus is worth zero if you can't get approved. We score approval difficulty from observed patterns by card — premium cards like the Amex Platinum score in the 40s (competitive, typically needs excellent credit), starter cards like Chase Freedom Unlimited score in the 90s. Spend Alignment checks whether bonus categories track real household spending — a card earning well only on airfare scores lower here than one earning well on dining and groceries, even with an identical signup bonus.

Key takeaway: offer strength is only 20% of the score — a headline-grabbing bonus on a card with a bad ROI, weak keeper value, or a category structure nobody's spending matches will still score "Average" or worse.

Elevated vs. near-high vs. tracked high vs. normal — what the labels actually mean

The badges you'll see on CreditPoints card pages map directly onto the historical-strength tiers above:

• "Highest Since Tracking" — equals or exceeds every offer we've observed for that card. Our strongest, most specific claim — used only once we have at least two prior tracked data points to compare against.
• "Near Tracked High" — within about 10% of that ceiling. Still a strong time to apply, but not necessarily the peak.
• "Elevated" — meaningfully above the card's typical offer, but not close to its tracked ceiling. The most common "good time to apply" badge you'll see.
• "Typical Offer" — running its normal, unremarkable range. No urgency either way.
• "Historically Weak" — below what we usually see, with no signs of a recent increase. Our explicit recommendation: consider waiting.

Most competitor sites collapse the top three into a single "elevated!" banner. We keep them separate — the gap between "near tracked high" and "elevated" can be the difference between a card worth applying for this week and one worth checking again next month.

Key takeaway: "elevated" and "tracked high" are not the same claim — elevated means above-typical, tracked high means the best we've ever verified for that specific card.

How long an elevated window typically lasts

There's no universal answer — it depends on the card and the competitive pressure on that product — but our own tracked history gives an honest illustration rather than a guess. Amex Gold's public offer moved from 60,000 Membership Rewards points (observed February 2026) to 75,000 (April 2026, ~7 weeks later) to 90,000 (May 2026, ~5 weeks later) to its current 100,000-point offer with an $8,000/6-month requirement as of August 2026 (~13 weeks later). Each step was a new tracked high at the time it was observed — meaning elevated windows in the offers we've tracked have run anywhere from about five to thirteen weeks before the next change. We don't extrapolate that onto cards we haven't tracked as closely, and we update the badge the moment a new snapshot changes the classification rather than leaving a stale "elevated" label up for months.

⚠️ Biggest mistake — Treating "elevated" as a countdown timer. We don't publish an expiration date for a badge because we don't have one — some elevated windows close in days, others run for a full quarter. Check back before applying if it's been more than a few weeks since you first saw the badge.

Key takeaway: elevated windows we've tracked have run roughly 5–13 weeks between changes on a single card — long enough to act on, but not something to assume will hold indefinitely.

Worked example: scoring the Amex Gold Card offer, August 2026

Here's the actual math, using Amex Gold's tracked data and CreditPoints' own realistic-tier assumptions.

The offer: 100,000 Membership Rewards points after $8,000 in purchases in 6 months, $325 annual fee, valued at roughly 2¢/point for a $2,000 bonus value.

1. Historical Offer Strength: 100,000 points is the highest offer we've tracked for this card — above the 90,000-point offer observed in May 2026, which was itself above the 75,000-point offer from April. Rank: tracked high → score 100 × 20% weight = 20.0 points.

2. First-Year ROI: Bonus value ($2,000) plus a realistic estimate of credit usage — call it $150 of the card's dining and Uber Cash credits actually redeemed in year one, a conservative fraction of what's technically available — for $2,150 gross value. Net of the $325 fee: $1,825. ROI = $1,825 ÷ $325 ≈ 5.6x, which lands in the "4x–6x" bracket: score 70 × 25% weight = 17.5 points.

3. Keeper Value: Strip out the $2,000 signup bonus from that $1,825 net figure and you're left at –$175 — meaning the realistic credit estimate alone doesn't fully offset the annual fee once the bonus is gone. That lands in a weak bracket: score 15 × 15% weight = 2.25 points. (This is intentionally conservative — it excludes ongoing 4x dining and grocery earnings, which a heavy spender would add back in.)

4. Approval Accessibility: Amex Gold sits in the moderate-competitive band — needs good to excellent credit, not the hardest approval in the lineup but not a starter card either: score 55 × 15% weight = 8.25 points.

5. Spend Alignment: 4x on dining and 4x on US groceries is one of the best category matches to typical household spending in the entire market: score 95 × 10% weight = 9.5 points.

Total: 20.0 + 17.5 + 2.25 + 8.25 + 9.5 = 57.5, rounds to 58 → "Solid."

Notice what happened: the historical-strength factor alone would suggest "Exceptional" — but the full score lands at "Solid," pulled down almost entirely by the conservative Keeper Value estimate. That's the score doing its job — a record-setting bonus on a card whose ongoing value depends on how aggressively you use its credits isn't automatically a top-tier card to hold forever, even though it's an excellent one to apply for right now.

Key takeaway: a tracked-high offer alone doesn't guarantee an Exceptional score — Amex Gold's August 2026 offer lands at "Solid" once first-year ROI, keeper value, approval difficulty, and spend fit are all weighed in.

Decision framework: when the score should actually change your behavior

Key takeaway: use the score's breakdown, not just the total — a "Solid" 58 driven by a real tracked high behaves differently than a "Solid" 58 driven by decent-but-unremarkable fundamentals across the board.

Editor's take

"Is this a good offer" is really three questions: is it strong for this card specifically, is it strong in dollar terms, and is the card worth keeping afterward. Most sites answer only the first, loosely. The CP Offer Score forces all three into one number so a record headline bonus on a mediocre keeper card doesn't get the same "apply now!" treatment as a genuinely well-rounded offer. Pair this with our minimum-spend strategy guide before chasing a specific bonus — a strong score on a card whose spend requirement you can't organically hit isn't a strong offer for you.

Disclosure: CreditPoints may receive compensation if you click through and are approved for cards mentioned in this article. We only recommend products we believe deliver genuine value to readers. See our editorial policy for details.