How We Test and Compare Social Media Tools
Our testing methodology for social media scheduling tools: what we buy, what we measure, how we score pricing, and the conflicts of interest we declare. Referenced by every comparison we publish.
Every comparison we publish points back to this page. Rather than repeat a methodology section in each post, the method lives in one place so you can audit it once and know what our conclusions are worth.
The conflict of interest, stated first
We build Social by InstantDM, a social media scheduler. It competes with most of the tools we write about.
That’s a real bias and I’m not going to pretend otherwise. What we do about it: name the conflict in every post where our product appears, include a section in each comparison explaining when the other tools are the better buy and mean it, run no affiliate links in comparison content, and publish the cases where we lose. If our tool lacks something the alternative has, it goes in the table.
Read our own recommendation sceptically. The rest of the analysis should stand up on its own.
What we actually do before writing a comparison
We buy the product. For self-serve tools that means signing up and paying list price on a normal plan. We avoid vendor-supplied press accounts, because those routinely have rate limits raised and features unlocked, and reviewing a product no customer can actually buy is worse than useless.
Then we connect real accounts and publish real posts, because a scheduler only gets interesting under load. We connect accounts across whichever platforms the tool claims to support, queue posts, and check what actually appeared, in what format, at what time. Claimed platform support and working platform support are not the same thing, especially for Instagram carousels, TikTok and YouTube.
We test the failure path rather than the happy path. Anyone can publish one text post. What we want to know is what happens when a token expires, when an image is the wrong aspect ratio, when a platform’s API rejects something, and whether the tool tells you it failed or just quietly doesn’t publish. Silent failure is the most expensive defect in this category and it’s the one least likely to show up in a demo.
We read the contract too. Renewal terms, how a seat is defined, what counts as a “social account”, how cancellation works. A lot of the real cost of these tools lives there rather than on the pricing page.
And we price a realistic team rather than one user. Headline prices are always quoted for a single seat, so we recompute for a one-person setup, a three-person team, and a ten-account agency. Per-seat and per-account models diverge violently at scale, and the headline number hides it.
Where we cannot test
Some platforms here don’t sell to individuals. Sprinklr and Brandwatch are sales-led, quote-only, and generally want annual contracts in the tens of thousands. We don’t have production access, and we say so in those posts rather than implying experience we lack.
For those tools we rely on vendor documentation, published pricing where any exists, procurement data from sources like Vendr and G2, and reporting we can link to. Where a figure is an estimate, we label it an estimate.
The seven things we score
Every tool gets assessed against the same criteria, though the weighting shifts depending on who a given comparison is for.
First, true cost at your team size, not the headline price, after per-seat charges, add-on accounts, and the annual-versus-monthly gap. Second, platform coverage that genuinely works, meaning which networks publish natively without a phone notification asking you to finish the post by hand. Direct Instagram carousel and TikTok video publishing are usually the dividing lines.
Third, publishing reliability: whether scheduled posts go out on time and whether failures are surfaced loudly. We weight this heavily, since it’s the entire job. Fourth, time to first post, because enterprise tools frequently need an onboarding call before you can do anything, and that’s a real cost even when nobody bills you for it.
Fifth, analytics you’d act on, meaning reporting that answers a question you’d change behaviour over rather than rendering follower counts attractively. Sixth, team and client workflow: approvals, roles, separate client workspaces, and whether adding a colleague triggers another full seat charge. Seventh, exit cost, because tools that make your scheduled queue unexportable are quietly expensive and nobody notices until they want to leave.
How we treat pricing figures
All prices come from vendors’ public pricing pages on the date stamped beneath each comparison table, in US dollars, for US billing.
Two traps we normalise for. Vendors nearly always display the discounted annual rate in large type and the monthly rate in small type, so we show both. And a $99 per-user tool isn’t comparable to a $79 per-account tool until you fix the team size, so we always state the assumption.
Pricing moves. If you’re reading this months after publication, check the vendor’s own page before buying.
When we change a verdict
We update posts when a vendor changes pricing, when a platform integration breaks or launches, or when a reader shows us we got something wrong. Updated posts carry a revised date at the top, and we correct in place and note material changes rather than quietly editing.
If you think a conclusion here is wrong, tell us and show your working. I’d rather fix it.
Watch these tools in action
A hands-on multi-tool test from someone with no product of their own to sell. Worth watching alongside anything we publish, for exactly that reason.
Frequently asked questions
Do you get paid to recommend tools?
No. We run no affiliate links in our comparison posts and we accept no payment for placement or ranking. Our only commercial interest is Social by InstantDM, which we build ourselves and declare in every post where it appears.
Do you actually pay for the tools you review?
For self-serve tools, yes — we buy a normal paid plan at list price rather than requesting a press account, because vendor-supplied accounts often have limits lifted. For enterprise platforms sold only through sales teams, we say so plainly and rely on published documentation and verifiable third-party reporting instead of pretending to hands-on access we do not have.
How often do you update pricing in your comparisons?
We re-check every price on a quarterly cycle and stamp each comparison table with the month it was verified. Social software pricing changes often, so we also tell readers to confirm on the vendor's own pricing page before buying.
Why do you sometimes recommend a competitor over your own product?
Because it is frequently the correct answer. A comparison that always concludes with our own tool is an advertisement, and readers work that out quickly. Where a rival genuinely fits a use case better — enterprise listening, franchise marketing, deep analytics — we say so.
What disqualifies a tool from being recommended?
Undisclosed pricing on a self-serve product, a support channel that never answers during testing, publishing failures we can reproduce, or contract terms that make cancellation unreasonably difficult.