How the 0-100 Privacy Score Is Actually Calculated

When people see a 0-100 score on their report, the first reaction is often "is this number even real". I get the skepticism. The score is computed on your own machine, but the logic is not a black box. You can pull it apart, and this piece does exactly that, so you know which kinds of content are riskiest and why two similar posts can land very different scores.

What kinds of risk make up the score

The tool scans your archive and mainly looks at four types of traces. Each one occupies its own band of danger:

Risk type	Common content	Why it is dangerous	Hard to fix
Contact info	phone numbers, emails, WeChat IDs	direct entry point; scammers and harassers reach you precisely	High
Location	home address, office, boarding passes, check-ins	tied to physical-world safety	High
Identity documents	ID photos, work badges, passport pages	can be abused to open accounts	Medium
Sensitive topics	rants about former employers, extreme opinions, health details	hurts job prospects and reputation	Medium

These four types are not weighted equally. The next sections explain why.

Why a phone number costs more points than a complaint

The difference comes down to actionability. A tweet with your phone number lets a bad actor call, text, and try credential stuffing with zero extra steps. A complaint about an old boss might surface during a background check, but only then. The first kind of harm is immediate and low-effort, so it carries more weight.

One detail people miss is repetition. Posting your number once three years ago is not the same as twenty consecutive posts with it. The risk is not twenty times larger, but it is clearly higher, because it shows you went a long time without noticing the problem.

Where the weights come from

Three sources cross-check each other. The first is which fields show up most often in public breach databases. The second is what victims report losing first in scam cases. The third is the density of user feedback that marks a post as the one that scared them most. When all three point the same way, that field gets a heavier weight. In other words, the weights follow real paths of harm, not guesses.

For Chinese users this matters because local scam patterns differ from the West. A WeChat ID in a tweet is a far more direct threat here than it would be for a US-only account, and the scoring reflects that. The report shows the scoring version and, for each flagged post, which category triggered it, so you are never told "this is risky" without a reason.

What the 0-100 number actually means

The score is only a summary. The useful part is the band:

• 0-30: high risk. Your archive likely contains hard info that locates or contacts you directly. Handle these first.
• 31-60: moderate. Scattered risk points, not dense.
• 61-80: good. Mostly soft risks from sensitive topics remain.
• 81-100: clean. Only historical traces left, low impact.

Most accounts land in the 30-60 band after a few years of casual posting, which is normal and not a cause for alarm. The bands are not decoration; they tell you which risk category to open first.

Does the score look at when you posted

Yes, with more weight on recent posts. An old tweet with your number still needs cleaning, but one posted last week is riskier because it is easier to find right now, so it carries slightly more weight. That is why cleaning early beats cleaning later: the longer a post stays, the more stubbornly it sits in the score. On the flip side, content you deleted three years ago is no longer counted when you re-run the check, and the score reflects that immediately.

How the privacy score differs from a credit score

People often mix this up with Alipay or credit scores. A credit score predicts whether you will repay debt, from borrowing and repayment history. A privacy score only measures how much hard information someone can dig out of your history, from your archive content. One is a financial prediction, the other is a risk-exposure snapshot. They share no data and should not. Raising your privacy score does not mean you are more creditworthy, only that you left fewer traceable traces behind. If a service ever asks for your privacy score, treat it as a red flag, because no legitimate lender needs it.

After clearing a category, run the check again rather than guessing. The report shows which posts were removed and how the band moved, so you see the step change instead of estimating it.

How to actually raise the score

Sort actions by payoff. Delete posts with phone numbers and addresses first, then ID photos, then sensitive topics, then stray location check-ins. After each type is gone, the score jumps a step. It does not climb evenly; it rises in clear steps once a risk category is cleared.

Here is a concrete example. Suppose your archive has 1,200 tweets. Twelve of them carry your old phone number from 2019, three show your home address, and about forty are angry posts about a previous employer. Deleting the twelve number posts moves the score the most, because contact info sits at the top of the weight table. Clearing the three address posts is the next biggest step. The forty rants barely move the number, which surprises people who expected them to matter most. The takeaway: spend effort where the weight is, not where the volume is.

Some people worry that a low score means the account is already compromised. A low score only means a large exposure surface, not that anyone has used the information. After one round of cleanup, the score usually drops, and most accounts land above 60.

If you have never uploaded an archive, read what a digital footprint check is for the overall flow, or upload it directly on the free check page to get your report in minutes. For deletion cost, see the pricing page. You can also reach the same tools from the homepage.

Every check on digital-footprint-health.shop runs on your device, and your data never leaves your computer. Pair it with privacy score versus credit score to see why these two numbers are not the same thing at all.