1b: Common mistakes
Common mistakes
1. Mixing up buckets and OR logic
Every bucket you create adds an AND requirement — the candidate must match something in each bucket. If you want a candidate to meet two separate criteria, use two buckets. If you want a candidate to match any one of several options, they all go in the same bucket.
Example — locations: If you're open to candidates in Stockholm, Gothenburg, or Malmö, add all three cities to the same bucket. If you put each city in its own bucket, TalentRiver will look for candidates who are in all three locations at once — which returns nothing.
Example — titles: If "Account Manager", "Sales Manager", and "Business Development Manager" are all acceptable, put them in one bucket. One of these is enough.
Example — skills: If you need someone with cloud experience AND backend language experience, use two buckets: one with AWS and Azure, one with Java and Kotlin. A candidate must have something from each bucket. If you put all four in the same bucket, someone with only AWS and no Java would match.
The rule is simple: same bucket means "any of these will do", separate buckets means "we need both".
2. Using keywords instead of skills
Keywords and skills look similar but behave very differently.
Keywords do a literal word match. TalentRiver looks for that exact word or phrase in the candidate's profile. This means a candidate who has "Java" written as "Java/JVM" or "core Java", or whose profile is written in Swedish, or who has a typo, may not appear. Keywords are useful for very specific terms — a niche tool name, an internal methodology, a certification acronym — where you want an exact match and nothing else.
Skills are far more flexible. TalentRiver searches across the whole profile and handles translations, synonyms, abbreviations, variations, and common misspellings. A search for Java as a skill will also catch candidates who list it as J2EE, JavaEE, Spring (implying Java), or have it written in another language. It understands context, not just text.
For almost everything — technologies, competencies, tools, functions — use skills rather than keywords. Use keywords only when you specifically need an exact string match and the skill search is returning too many loosely related results.
3. Adding non-title terms into the title field
The title field is built for job titles only. When you add something that is not a title — a skill, a keyword, a function, a level like "senior" — you lose the subcriteria that make the title field useful.
When you use the title field correctly, TalentRiver lets you control how long someone has held that title and when. You can set "Account Manager, now, for 2–5 years" and get candidates who currently hold that title and have done so for a meaningful period. The moment you put something non-title into that field, those subcriteria no longer apply in a meaningful way.
What to do instead:
If it is a skill or technology, add it as a skill criterion.
If it is a seniority level like "Senior" or "Lead", use the experience level settings within the title or Relevant Work Experience criterion.
If it is a function or department, use Relevant Work Experience and set the function there.
Keeping the title field clean is one of the easiest ways to make sure your subcriteria — timing, duration, current vs. past — work as intended.