Retirement Planning Shouldn't Feel Like Homework

Nobody sits down on a Saturday wanting to read fifteen articles about Medicare, RMDs, and Social Security timing. They want one answer to their actual situation — and that gap is the entire reason SavedNest exists.

Someone might be trying to understand Medicare. Someone else might be wondering how retirement income is taxed. Another person may be trying to understand an inherited IRA, long-term-care costs, benefits eligibility, or what a particular state does differently.

Those are not really isolated questions.

They are connected questions that appear at different points in the same larger decision journey.

That realization changed the way I thought about SavedNest.

A retirement website doesn't become more useful simply because it contains more retirement information. It becomes more useful when the information is organized around the decisions people are actually trying to make.

What I was really building

At first glance, SavedNest can look like a retirement information site. But the deeper project became much more interesting: build a system that helps people understand benefits, retirement income, taxes, long-term care, and related financial decisions without forcing every visitor to start from scratch.

That meant building more than articles.

The current site illustrates that direction particularly clearly. SavedNest now includes tools such as a Benefits Eligibility Checker that brings several programs into one screening flow, while its editorial system connects national explanations with state-specific guides. That is a very different architecture from a simple chronological blog.

The first lesson: retirement questions are rarely single questions

One of the biggest things I learned while working on SavedNest was how quickly a seemingly simple question expands.

Consider a person asking whether a retirement benefit is available to them.

The next questions may be:

• What income counts?
• Does my state use a different rule?
• Does Medicare change anything?
• What happens to my spouse?
• What happens to my savings?
• What if I need long-term care later?
• Is there a calculator or eligibility check I can use?
• What should I do next?

If every one of those questions is treated as an unrelated article, the visitor has to build the mental model themselves.

That is where information architecture becomes part of the product.

From content library to decision system

The architecture I increasingly wanted was less like a bookshelf and more like a map.

The visitor doesn't necessarily need to see that structure explicitly. But the site should understand it.

That idea also explains why state-specific content matters so much on SavedNest. A national explanation can introduce a concept, but the actual answer may depend on the state. The current SavedNest Medicaid work makes that distinction explicit: its state guides emphasize that income limits, asset rules, waiver programs, and other details vary significantly by state.

Then the search data started telling me something

This is where the project became an experiment rather than just a build.

The Search Console snapshot I captured for SavedNest showed 3.12K impressions over the selected three-month period, but only 6 clicks. Average CTR was 0.2%, with an average position of 70.8.

That isn't a traffic success story yet. And I don't want to pretend it is.

But it is useful evidence.

The more interesting part of the screenshot was the query mix. SavedNest was receiving impressions for searches including Medicare Savings Program, taxes on retirement income calculator, inherited Roth IRA RMD rules, retirement tax calculator, and Arkansas Medicaid eligibility requirements for seniors.

Those queries are telling me that Google is already associating the site with several different decision areas — Medicare, retirement taxation, inherited retirement accounts, and state-specific benefits.

The next question isn't simply, "How do I get more impressions?"

It's:

What should the site do when someone arrives through one of those questions?

Search Console became a product-development signal

This is one of the biggest lessons I took from SavedNest.

Search data can be treated as a report card: impressions went up, clicks went down, rankings changed.

Or it can be treated as a discovery tool.

If people are finding a site through combinations of questions that I didn't originally imagine as one cluster, that is architectural information.

For example, a query about inherited IRA rules can sit near retirement income planning. A Medicare Savings Program query can lead into benefits eligibility. A long-term-care question can lead into Medicaid, insurance, veterans' benefits, personal savings, and family contribution.

The content starts behaving like a network.

The second lesson: more pages don't automatically mean more usefulness

SavedNest has grown substantially. But the experience taught me to be careful about equating scale with quality.

A site can have hundreds of pages and still make the visitor work too hard.

The better question is whether the pages reinforce one another.

That distinction matters especially in a subject like retirement planning, where one answer can create three more questions.

Building tools changed the role of the content

Another shift was the relationship between editorial pages and interactive tools.

A guide can explain what a program is. A tool can help a visitor organize the information around their own situation.

That makes the tool more than a feature.

It becomes a destination in the information architecture.

SavedNest's Benefits Eligibility Checker is a good example of that direction. The current tool asks for a small set of household inputs and checks several programs together, while explicitly explaining that the result is an informational estimate rather than an official determination.

That creates a useful progression:

That is much closer to how I want digital products to behave.

The third lesson: state-level detail makes architecture unavoidable

Retirement and benefits content contains a structural problem that is easy to underestimate: the same phrase can mean something different depending on where the person lives.

SavedNest's current state-level Medicaid work made that impossible to ignore.

For example, its Connecticut guide explains a 2026 asset limit of $1,600 for a single long-term-care Medicaid applicant and distinguishes the state's home/community-care pathway from nursing-home Medicaid. Its California guide describes a very different 2026 asset-limit structure and a scheduled 2027 change.

The point isn't that one state's rules are more important than another's.

The point is that a national article alone cannot carry the whole experience.

The architecture has to tell the visitor when the national explanation ends and when local rules begin.

What building SavedNest taught me about trust

When a site deals with retirement, Medicare, Medicaid, taxes, or benefits, usefulness isn't enough. The structure has to communicate caution.

I've become much more deliberate about distinguishing:

• general explanations from official determinations;
• national rules from state-specific rules;
• current figures from historical figures;
• planning observations from authoritative requirements;
• educational tools from professional or government decisions.

That is not merely legal or editorial housekeeping.

It changes the design.

The current SavedNest pages, for example, explicitly tell readers when figures are state-specific and direct them to the relevant agency or professional for confirmation.

And then there was the indexing problem

The Search Console screenshot showed another part of the real build: not every page that exists on the site is indexed.

At the time of the snapshot, Search Console showed 127 indexed pages and 96 not indexed.

The largest exclusion category shown was Discovered – currently not indexed, with 73 pages. There were also smaller groups such as pages with redirects, redirect errors, 404s, noindex exclusions, alternate pages with proper canonicals, and a server error.

That was a useful reminder that building a page and getting a page into Google's searchable index are different events.

It also reinforced something I have learned across projects: technical SEO isn't a final checklist that gets applied after the site is finished. It is feedback about the architecture itself.

So what am I actually learning from SavedNest?

Probably more than I expected when I started.

Where SavedNest goes from here

The next phase isn't simply "write more."

It is about making the existing system easier to move through.

I want a visitor to be able to enter through almost any meaningful question and naturally find a useful next step:

That is the bigger experiment.

Can a website stop behaving like a collection of pages and start behaving like a useful decision-support system?

SavedNest is one of the projects where I'm trying to find out.