Product blog

Product blog

Product updates, field guides, and field notes—each with links to docs and the hub where relevant. High-level milestones also appear on the changelog.

  1. A decision funnel sending an idea through three evidence gates toward build, pivot, or drop outcomes

    Field guide · Jun 22, 2026

    Validate a business idea before you spend a dollar

    A cheapest-first sequence for proving demand before you build anything, so you spend money on confirmed interest rather than hope.

    • Real signal vs vanity signal
    • The mom-test way to interview
    • Set a kill threshold before you test

    Novus Stream Solutions (hub)

    Read
  2. A decision tree branching from a root question into three ecommerce platform archetypes: hosted, open-source, and headless, each tagged with a cost shape.

    Field guide · Jun 22, 2026

    Choosing an ecommerce platform: hosted vs open-source vs headless

    A decision framework for matching an ecommerce platform to your real constraints, not the hype around any one product.

    • Compare on total cost, not sticker price
    • Most small sellers should start hosted
    • The specific reasons to deviate

    Novus Stream Solutions (hub)

    Read
  3. Two render paths side by side: a single-page app downloading a large JavaScript bundle to a blank screen before showing content, versus a static-first path that ships HTML the browser paints immediately and enhances afterwards

    Field guide · Jun 18, 2026

    Static-first: when a small site doesn't need a single-page app

    A plain-English case for building static-first instead of reaching for a single-page app by reflex: what a SPA actually costs in load time, complexity, and fragility; why most small sites are content with a little interactivity, not apps; what progressive enhancement buys you; how to decide per-feature rather than per-site; and the narrow cases where a SPA genuinely earns its keep.

    • Ship HTML first, enhance second
    • SPAs cost load time and complexity
    • Decide per feature, not per site

    Novus Stream Solutions (hub)

    Read
  4. A content library shown as a garden of pages: thin and duplicate pages being pruned, two competing pages merging into one stronger page through a 301 redirect, and the remaining pages rising in rank

    Field guide · Jun 18, 2026

    Pruning and consolidating content without losing rankings

    A careful playbook for cleaning up a content library: why thin and duplicate pages drag down the whole site, how to audit what you have by traffic and intent, the keep / refresh / merge / remove decision, how keyword cannibalization happens and how to fix it by consolidating, and why the 301 redirect is the tool that lets you do all of this without losing the equity you built.

    • Merge competing pages, redirect the rest
    • Fix keyword cannibalization
    • 301s preserve the equity you keep

    Novus Stream Solutions (hub)

    Read
  5. A scale balancing a small stack of monthly running costs against a rising bar of ad revenue, with a break-even line marking the monthly visitor count where the two meet

    Field guide · Jun 17, 2026

    How much traffic a free tool needs to pay for itself

    A clear-eyed estimate of the traffic a free, ad-supported tool needs to fund itself: what RPM really means and why it is the number that matters, the genuinely small cost base of a serverless tool, the simple break-even math, why a new tool earns far less per visit at first, and why traffic quality changes the answer as much as quantity. Educational, not financial advice.

    • Break-even = costs ÷ RPM
    • Serverless costs are tiny
    • New tools earn less per visit at first

    Novus Stream Solutions (hub)

    Read
  6. A pipeline taking raw images through stages — generate a draft alt text, rename to a meaningful filename, strip EXIF, human review — and out to a published image with complete, clean metadata

    Field guide · Jun 17, 2026

    Automating alt text and image metadata at scale

    A practical approach to image metadata when there are too many images to handle by hand: what metadata actually covers, why alt text is for people before search engines, how to use AI to draft alt text without shipping confident nonsense, generating filenames and stripping EXIF as a pipeline step, and a build-time check that keeps the whole library honest.

    • Automate the toil, keep the judgement
    • AI drafts alt text; a human approves
    • Strip EXIF; check it at build

    NSS Background Remover

    Read
  7. Three policy documents — privacy policy, terms of service, refund policy — each with a checkmark, sitting on a foundation labelled honest and matching your actual practice

    Field guide · Jun 17, 2026

    Terms, privacy, and refund policies without a lawyer

    A practical, non-lawyer guide to the three policy pages a small online business actually needs: what a privacy policy, terms of service, and refund policy each do, why starting from a reputable template beats a blank page, why the words must match what you actually do, the case for plain language over legalese, and the specific situations where you should stop and hire a professional. Educational, not legal advice.

    • Three pages: privacy, terms, refunds
    • Match the words to your behaviour
    • Know when to hire a lawyer

    Novus Stream Solutions (hub)

    Read
  8. One large app bundle split into a small initial shell plus many per-route chunks that load only when a tool is opened, with the AI model deferred until first use

    Field guide · Jun 16, 2026

    Code-splitting a large web app: how lazy routes keep it fast

    How the NSS Background Remover stays fast to open despite many tools: route-level code-splitting so each tool is its own chunk, lazy boundaries that defer the heavy parts, deferring multi-megabyte AI models until a tool actually needs them, and the measurements that tell you whether any of it worked.

    • Each tool is its own chunk
    • Models load only on demand
    • Measure first-load JS, not total

    NSS Background Remover

    Read
  9. A tool’s first screen designed to teach — a single clear drop-zone action, a sample to try, and a one-line hint — turning an empty state into onboarding without documentation

    Field guide · Jun 16, 2026

    The first-run experience: onboarding a tool nobody reads docs for

    How to onboard a tool that no one reads docs for: why the first screen is the only instruction most users will ever see, designing the empty state to teach rather than just sit there, making the single most important action unmissable, getting to a first result fast, and layering help so it is there without being in the way.

    • The empty screen is the manual
    • One obvious first action
    • A fast, visible first win

    NSS Background Remover

    Read
  10. A twelve-month revenue curve with a Q4 peak and a January and summer trough, overlaid with a flat budget line set to the average and a buffer band that fills in peaks and drains in troughs

    Field guide · Jun 16, 2026

    Planning cash flow around seasonal ad revenue

    A clear-eyed look at the seasonality of ad income and how to plan around it: why ad rates swing by season regardless of your traffic, the danger of treating a Q4 peak as your baseline, budgeting to a trailing average rather than the high, building a buffer that smooths the trough, and separating the revenue cycle from your spending decisions. Educational, not financial advice.

    • Ad rates peak in Q4, slump after
    • Budget to the average, not the peak
    • A buffer smooths the trough

    Novus Stream Solutions (hub)

    Read