Filter first, buy last: Sandi Hermawan’s expired-domain workflow
By Lucy Kim, founder of Karma.Domains
Sandi Hermawan’s Indonesian walkthrough, “Cara Mencari Expired Domain Berkualitas dengan Karma.Domains”, spends more than twenty minutes inside a real domain-research session. He moves from the public database to filters, reports, auctions, backorders, buy-now listings, and a check of a domain he already owns.
The video is detailed, but the method underneath it is compact. Find candidates, filter them, compare SEO metrics, read the archive, check risk signals, and only then consider buying. That order is the useful part.
Turn a live database into a small candidate set
The walkthrough opens with the scale of the database and the number of available filters. Live totals change, so the durable takeaway isn’t a particular domain count. It is that no one should review the inventory row by row.
Sandi shows more than 90 filter options across acquisition type, marketplace, price, Wayback Machine history, authority, backlinks, anchors, traffic, Ahrefs, Semrush, Majestic, Moz, Keywords Everywhere, Google, and Similarweb data. He also points out natural-language AI search, which can translate a plain request into an editable set of conditions.
The filters need a brief. “Find good domains” is too vague. “Show recently added .org domains in education, with an established archive, relevant backlinks, and no toxic-topic history” gives the search a job it can actually perform.
Use metrics to raise questions
Sandi opens individual reports and compares Karma metrics with data associated with Moz, Ahrefs, Semrush, Majestic, Google, and Similarweb. He looks at authority, backlink counts, referring domains, indexed pages, and traffic by country.
That side-by-side view is useful when the numbers disagree. A strong score from one provider and weak evidence elsewhere is a reason to inspect further. So is a high backlink count paired with few meaningful referring domains, or historical traffic from a country and language unrelated to the site’s former topic.
The video sometimes reacts enthusiastically to individual scores. I would treat those scores as screening signals. A domain earns attention when several independent clues make sense together.
Read the archive as a timeline
The most valuable part of the walkthrough is the move from the summary metrics to the old site itself. Sandi checks the last screenshot, titles, descriptions, language, category, archived snapshots, redirects, content changes, toxic topics, and server responses.
That turns a domain report into a timeline. If the domain was a Russian-language education site for years, the language, categories, backlinks, and archive should support that story. If it later became a parked page or changed to an unrelated commercial topic, that transition belongs in the buying decision.
The archive also helps explain why a metric exists. A healthy backlink profile attached to a long-running organization is different from a similar score created during a short period of aggressive link building.
Check the acquisition route after the history
Sandi demonstrates auction, backorder, buy-now, and freely registrable expired-domain sections. He follows marketplace links and looks at bids, prices, and time remaining.
Get Karma Domains’s stories in your inbox
Join Medium for free to get updates from this writer.
This is where urgency can distort judgment. A countdown timer makes an average domain feel scarce. A low opening bid makes a questionable name feel cheap. Put the domain through the same history and backlink checks before opening the marketplace, then set a maximum price based on the intended use.
The domain report and the marketplace answer different questions. The report helps you decide whether the asset deserves further attention. The marketplace tells you whether it is still available and what acquisition may cost.
Run the same checks on domains you already have
Later in the video, Sandi uses the domain-check feature on one of his own sites. This is a useful extension of the workflow. Due diligence is not limited to domains currently listed in the database.
An existing portfolio can be checked for archive continuity, spam indicators, authority changes, backlink patterns, indexing, and category drift. An agency can use the same process before accepting a domain from a client or before recommending a redirect. A domain investor can use it to document an asset before listing it for sale.
The objective remains the same: assemble enough evidence to understand what the domain has been, not merely what its current homepage says.
Sandi’s buying rule
Near the end, Sandi compresses the entire tutorial into a sequence: search, filter, check SEO metrics, inspect Wayback Machine history, review Karma Score, look at toxic topics and redirects, then check backlinks manually.
He follows it with a short warning: “Jangan dibalik ya, jangan lihat domain langsung beli”. In English: don’t reverse the order and buy as soon as you see the domain.
That warning is easy to remember because auction interfaces encourage exactly the opposite behavior. They show the name, the clock, and the price first. Research has to put history back in front of urgency.
A six-step version for daily work
Here is the practical workflow shown across the video:
- Define the intended use, topic, language, extension, and budget.
- Search the relevant inventory and apply a narrow first-pass filter.
- Compare authority, backlinks, referring domains, traffic, and indexing across sources.
- Review archived content, topic continuity, language changes, redirects, and toxic themes.
- Check the current marketplace state and set a price ceiling.
- Manually validate the finalists before registering, bidding, or placing a backorder.
This is also why a platform such as Karma.Domains for expired-domain research is most useful before the final buying step. It gathers the evidence and reduces the list. It does not turn a high score into a guarantee.
Sandi closes with another concise observation: “Karma domain bukan sekedar alat untuk mencari expired domain”. His tutorial proves the point. Discovery fills the top of the funnel; the real work is deciding which names should leave it.
