Customer tension A
More hints can create more doubt.
When suggestions arrive without enough explanation, customers must do the work of separating useful evidence from plausible noise.
Independent product design leadership exploration · Ancestry
Every discovery should build on the last.
A strategic product concept for making family-history research feel less like moving between tools—and more like making steady progress on one meaningful question. The opportunity isn't another feature. It's a more intelligent, continuous research experience.
Research archive · case 014Family history
Eleanor Whitmore
1892—1974 · LEEDS, ENGLAND
7 SOURCES ATTACHED2 QUESTIONS OPEN
Research thread · activeOne question. Clear next steps.
Evidence, people and unresolved details stay connected as the research evolves.
The central thesis
Ancestry can create more value by helping people understand what they know, what they don't know yet, why a clue matters, and what to do next—without losing context as they move through the product.
01 / Strategic framing
Design around the researcher's progress—not the product's navigation.
Tree, records, profiles and comparison tools each solve important jobs. The leadership opportunity is to make them behave like parts of one coherent journey.
Start with the customer’s unfinished question.
A person isn't thinking, “I need to use the records feature.” They're thinking, “Is this the right Eleanor? Where did she live before Leeds? Does this record actually support the relationship I have in my tree?”
That distinction changes the design brief. Instead of optimizing isolated screens, design the connective tissue that preserves intent and evidence across the journey. The product should remember the research problem—not just the last page visited.
This is an opportunity hypothesis, not a claim that Ancestry lacks the underlying capabilities. The concept builds on familiar genealogy workflows and explores how their combined experience could become more coherent, transparent and confidence-building.
01
Carry context forward
Keep the selected person, research question, relevant relationships and prior decisions available as customers move between surfaces.
02
Make evidence legible
Show what supports a conclusion, what conflicts with it, and what remains an assumption. Never turn a relevance score into false certainty.
03
Recommend a meaningful next step
Suggest the action most likely to resolve the current question—not simply the next feature to visit.
04
Protect the human story
Make complex research feel approachable while keeping source history, relationships and user control intact.
02 / Critical thinking
The hard problem is not discovery. It's knowing what to trust.
Genealogy is a domain where a fast answer can be worse than no answer. The experience must balance momentum with historical accuracy and human judgment.
Customer tension B
One person can look like several people.
Names, dates and places vary across records. Similarity alone is not proof—and an incorrect merge can damage years of research.
Design opportunity
Turn activity into understanding.
Connect each clue to the question it may answer, make conflicts visible, and let customers decide when the evidence is strong enough.
03 / Experience concept
A research workspace that keeps the thread intact.
Explore the prototype tabs. The sample is deliberately focused on one research case so the connective experience—not a broad feature catalog—is the thing being evaluated.
ancestry RESEARCH THREAD · CONCEPT
Research workspace · Eleanor Whitmore
Research brief · Active investigation
Who was Eleanor before Leeds?
Understand whether the records point to one continuous life—and what evidence is still missing.
Research question stays with youOpened from Eleanor's profile · 7 sources reviewed · Last decision: keep two identities separate
What we can support
Leeds is a consistent anchor.
Two independent records place an Eleanor Whitmore in Leeds during the period under review. The dates are compatible with the current profile.
What remains uncertain
The earlier surname is unresolved.
A separate Eleanor Wells has a similar age and location. The available records do not yet establish that they are the same person.
Why it matters
A wrong merge changes the family story.
Combining the profiles now could attach unrelated parents and sources to Eleanor's timeline.
Recommended next action
Find an independent link.
Prioritize a marriage, household or address record that connects the names—not another record that merely repeats the same dates.
Evidence ledger · Source provenance
What supports this story?
Separate observations from interpretations. A source can support one detail without proving an entire identity.
Evidence standardIndependent sources matter more than repeated copies of the same underlying record.
RECORD / OBSERVATIONASSESSMENTWHAT IT SUPPORTS
1901 census indexLeeds · age 9 · Eleanor Whitmore
Consistent
Early timelineSupports a Whitmore household in Leeds.
1911 census transcriptLeeds · age 19 · Eleanor Whitmore
Consistent
ContinuityAge and place align with the current profile.
Marriage indexLeeds · Eleanor Wells · 1913
Needs review
Possible name linkPotentially relevant; spouse and certificate details need confirmation.
Parish register extractSame district · different parent names
Conflict
Identity not provenCould refer to another person with a similar name.
Relationship context · Identity graph
People are connected by evidence—not just lines.
Relationships should reveal why a record matters and where an assumption could affect the wider tree.
Current profile
Eleanor Whitmore
Born 1892 · Leeds, England. Seven sources attached. Current working identity retained.
Possible related identity
Eleanor Wells
Born about 1893 · Leeds, England. Two sources. Similarity is a lead, not a confirmed relationship.
Known relationship
Thomas Whitmore
Spouse relationship appears in a later source. Earlier records need to be checked for continuity.
Potential impact
Two branches may be affected.
Any identity decision could change parent, spouse and descendant connections. Preview the impact before applying it.
Decision support · Human in control
Should these two profiles be combined?
The responsible answer today is no—not until an independent source establishes the connection.
COMPARISONASSESSMENTDECISION IMPLICATION
Birth year1892 ↔ about 1893
Compatible
Weak signalAge proximity alone cannot establish identity.
BirthplaceLeeds ↔ Leeds
Compatible
Weak signalShared location is useful context, not proof.
SpouseThomas Whitmore ↔ not confirmed
Unresolved
Critical gapFind a primary or independent source linking the relationship.
Parent detailsConflicting names in parish extract
Conflict
Do not mergePreserve both histories until the discrepancy is understood.
Decision saved in this prototype only. No tree data has been changed.
INTERACTIVE CONCEPT · Fictional sample people and records · Buttons demonstrate navigation and local prototype feedback only. No real Ancestry data or services are connected.
04 / Product decisions
Make the system useful without making it overconfident.
The concept is intentionally opinionated about trust, explainability and control. These are product requirements, not decorative principles.
What I would protect at all costs.
Family history is personal. A system that feels fast but quietly damages a person's research will erode trust. The experience must make uncertainty visible and keep the customer in charge.
No opaque scores
Don't present a single percentage as truth. Explain which attributes align, which conflict, how sources relate, and what evidence is missing.
Preserve provenance
Every conclusion should be traceable to its source. Distinguish original records from indexes, derivatives and repeated copies.
Make uncertainty actionable
Show the next best research action and why it could resolve the question. Don't leave customers with a vague “needs review” label.
Preview consequences
Before a merge or relationship change, show affected people, attached sources, alternate names and what will happen to each record.
Respect customer intent
Let customers save a research question, pause it, dismiss a suggestion or return later without losing their place or being pushed by notifications.
05 / How I would lead the work
Prove the connective value before scaling the system.
A focused sequence gives product, design, engineering and data partners a shared path from customer problem to measurable outcome.
Find the highest-friction journey.
Establish where context breaks and which customer question has the greatest unmet need.
- Map end-to-end research journeys
- Review behavioral and support signals
- Interview novice and experienced researchers
- Baseline trust, task success and drop-off
Test the smallest useful thread.
Prototype one connected research task before committing to a broad platform redesign.
- Choose one high-value identity question
- Test context persistence across surfaces
- Evaluate evidence explanations
- Test safe conflict and merge decisions
Build a repeatable experience system.
Turn validated patterns into shared principles, components and team practices.
- Define shared research-context model
- Align product and engineering roadmaps
- Instrument outcomes and guardrails
- Coach teams through adoption and iteration
06 / Measurement strategy
Measure progress, confidence and quality together.
Engagement is an outcome—not the only outcome. A healthy experience helps people make meaningful progress without encouraging careless decisions or empty activity.
Research task completion
Share of customers who resolve or make a documented next-step decision on their stated research question.
Primary outcome · effectivenessContext continuity
Reduction in repeated searches, lost state and abandoned transitions while moving between relevant surfaces.
Journey quality · frictionEvidence comprehension
Can customers correctly explain why a record supports—or does not support—a proposed conclusion?
Trust · understandingDecision quality
Precision of accepted identity resolutions, correction rates and reversals after merges or relationship changes.
Guardrail · accuracyResearch confidence
Customer-reported confidence calibrated against evidence quality—not simply confidence that the interface feels easy.
Trust · calibrated confidenceMeaningful return
Return sessions that advance an open question or build on prior work, not raw visit frequency alone.
Long-term value · retention07 / Leadership point of view
Set a clear direction. Give teams room to do their best work.
The role is not to be the person with every answer or the most polished screen. It is to align people around the right problem and raise the quality of decisions across the organization.
My approach to design leadership.
I would bring product, design, engineering, data and customer-facing teams around a shared view of the customer journey. Make the problem concrete, agree on the principles that matter, and create enough clarity for teams to move with confidence.
That means being strategic about where to invest, practical about what can be learned quickly, and hands-on when the work needs a clear creative or experience direction. It also means coaching designers to connect craft to customer and business outcomes—not just deliverables.
Direction over decorationTranslate business priorities and customer insight into a memorable experience vision teams can use to make decisions.
Alignment without silosBuild shared ownership across product, engineering, data and partner teams around end-to-end outcomes.
High standards, low egoRaise the bar for clarity, craft and evidence while creating space for debate, learning and designer growth.
Learn, measure, adaptUse research and behavioral signals to challenge assumptions, make trade-offs and evolve the direction.
The opportunity
Make every discovery feel like progress.
The strongest Engage experience won't just help customers find more. It will help them understand more, make better decisions and feel that the work they've already done is carrying them forward. That's the product opportunity—and the kind of cross-functional design challenge I would be excited to lead.
Independent speculative concept. This is an original, self-directed portfolio exploration. It is not commissioned, endorsed or reviewed by Ancestry and does not represent Ancestry's internal strategy, product roadmap, systems or performance. The people, dates, records and relationships in the prototype are fictional sample data. The archival portrait is used illustratively and is not a photograph of the fictional person. Product opportunities and measurement ideas are hypotheses for research and validation, not claims about current product deficiencies or measured results.