Ask five people how to improve domain authority and you'll get answers ranging from "publish daily" to "buy expired domains" — most of it noise. Domain Authority is Moz's estimate of how well a site might rank, calculated almost entirely from your backlink profile. Understanding that single fact cuts through the myths and tells you exactly where to spend effort.
This guide explains what actually moves DA-style scores, what merely feels like it should, and how to raise the number without resorting to tactics that put your real rankings at risk.
Quick Answer: Domain Authority rises when you earn more dofollow links from many unique, strong referring domains — and essentially nothing else moves it. Content quality and traffic growth help indirectly by attracting those links.
What DA Actually Measures (and What It Doesn't)
Moz's Domain Authority is a comparative score from 1–100, machine-learned against their index of links. Three properties follow from that definition:
- It's logarithmic. Moving 20→30 is far harder than 10→20, because you're out-climbing sites with massive link profiles.
- It's relative. Your score can drop while your link count grows, if the wider web grows faster.
- It's not a Google metric. Google uses its own hundreds of signals; DA approximates one family of them.
The Only Direct Lever: Referring Domains
DA weights links from unique root domains heavily. Fifty links from one site move the needle less than ten links from ten independent, well-linked sites. That shapes every tactic worth doing:
- Digital PR and newsworthy assets. Original surveys, data studies and industry reports attract citations from publications that link to few other sites in your space — the highest-leverage placements available.
- Guest contributions on real editorial sites. Not guest-post networks — actual columns and expert commentary where editors vet contributors.
- Linkable resource building. Free tools, definitive guides and templates accumulate passive citations year after year.
- Unlinked brand mention reclamation. Sites that already mention you are the warmest outreach list you'll ever have.
- Broken-link and replacement outreach. Offering your page as a substitute for dead resources converts existing citation habits into new referring domains.
What People Believe Moves DA — But Doesn't
| Action | Effect on DA | Why It Persists as a Myth |
|---|---|---|
| Publishing more blog posts | No direct effect | Content earns links over time, so correlation looks like causation |
| Improving site speed | No direct effect | Better UX attracts shares and links indirectly |
| Social media activity | No direct effect | Viral posts generate the links that do move DA |
| Buying expired high-DA domains and redirecting them | Tempting short-term bump | Moz recrawls and devalues dropped spam; Google may penalise manipulative redirects |
| Internal linking improvements | No effect on DA itself | Helps real rankings, which people conflate with the metric |
Realistic Timelines and Expectations
New sites typically sit in the single digits for months. Consistent earning of relevant editorial links usually produces visible movement — a few points per quarter early on — with each subsequent point getting harder as you climb the logarithmic scale. Sites stuck below 20 almost always have a volume problem (too few referring domains); sites stuck between 30 and 50 usually have a quality problem (links concentrated on low-authority sources).
A Word on Shortcuts
Link farms, PBNs and bulk link packages can inflate third-party metrics quickly — that's precisely why they sell. The trade: money spent on links Google later ignores or penalises, a profile pattern that flags future manual reviews, and cleanup costs that dwarf the original invoice. Metrics that exist to predict rankings shouldn't be pursued in ways that destroy actual rankings.
Key Takeaways
- DA is a logarithmic, comparative estimate built from links — not a Google input.
- Unique referring domains matter more than raw link counts.
- Content, speed and social help only insofar as they attract links.
- Chasing the metric with purchased networks risks the rankings the metric exists to predict.