Eleven publicly marketed SD-WAN and SASE offers — Cisco, Fortinet, Palo Alto Networks, HPE Aruba / Juniper, Arista VeloCloud, Versa, Zscaler, Netskope, Cato Networks, Cloudflare and Check Point — benchmarked, correlated, compared, heat-mapped, dependency-mapped and gap-analysed from the 20 November 2026 landscape analysis, with an illustrative global cloud PoP footprint.
Cisco ranks 1 of 11 on the SASE-weighted model (4.30 / 5), but the spread at the top is small — choose by category and estate, not by the total alone.
The document weights integrated security (20%) and SASE convergence (15%) most heavily. Re-weight for your buyer profile and watch ranking, radar and bubbles move.
Weighted mean of nine criteria (document default = 4.30 for Cisco).
Toggle vendors to overlay nine-dimension profiles.
D3 pack layout. Every bubble is one vendor × criterion; area grows with the score. Regroup by vendor or by criterion.
Head-to-head on the nine scored criteria, nested treemap of the scorebook, the offers table from the analysis, and share-of-total pies.
Pick any two vendors. Bars extend from the centre.
Criterion family → criterion → vendor; tile area = score.
SD-WAN, SASE / SSE and supporting components from section 4 of the analysis. Filter by any keyword.
Category mix, share of weighted score, or published / illustrative PoP counts.
Do vendors that are strong on SD-WAN also score on operations? Pearson r across eleven vendors (n=11 — directional, not statistically conclusive).
Pearson r of the 1–5 score vectors.
Choose two criteria. Overlapping vendors are fanned out.
Ribbon width = vendors scoring ≥ 4 on both criteria.
Ribbon width = Σ of the lower score per criterion (shared depth).
Benchmark scores, feature-matrix ratings (Strong / Good / Partial / Limited), and the five-axis lock-in profile from the analysis.
Vendors × nine weighted criteria (1 = weak · 5 = leading). Right: weighted total on the document default.
Section 5.1 ratings recoded Strong=4, Good=3, Partial=2, Limited=1.
5 − score. Hotter = further from a leading native capability.
Section 8.3: Low / Medium / High switching-cost axes.
A production SASE design is a stack: underlay → edge → overlay → SSE → identity → observability → orchestration. A failure lower down propagates up.
Vendors (left arc) link to the layers they depend on. Hover a node to isolate connections.
Cisco’s layered offers, then every vendor with headline strength (mint) and gap (rose). Drag nodes.
Cisco’s material gaps are console unification, dual-platform coherence, SSE depth versus specialists, and no owned private backbone — not SD-WAN routing.
Stacked 5 − score. Shorter bar = closer to a leading native score of 5.
Average gap by criterion across all eleven vendors.
Follow a buying context to its likely leaders, then to the gap each leader still needs to prove in a PoC.
Buyer-profile → shortlist mapping is from section 6.4. Vendor → gap mapping is taken from sections 7.1 and 7.2.
Buyer profile → shortlisted vendor → the decision test to run. Hover to read the path.
A Voronoi partition of major internet hubs: each cell belongs to the nearest mapped PoP — a geographic reading of published footprint scale, not an official site list.
Toggle vendors to recompute the partition. Scroll to zoom.
Added context — city points are illustrative. Hub coverage is scaled to published PoP / data-centre counts (Cloudflare 330+, Zscaler 200+, Fortinet 170+, Palo Alto 150+, Netskope 125+, Cato 85+, Check Point 80+, Cisco Umbrella ~47). Versa, HPE and Arista have no single published count in the source analysis; their dots follow the qualitative “own / partner PoP” language and are labelled illustrative. Not an official inventory. Verify before customer use.
Public counts where known; qualitative for Versa, HPE, Arista.
Share of each vendor’s illustrative hubs by macro-region.
How each offer was assembled — Viptela, CloudGenix, Silver Peak, VeloCloud, Infiniwan, Perimeter 81 — plus the four-phase evaluation roadmap from the analysis. Scroll to zoom, drag to pan.
One lane per vendor; stems mark milestones.
Acquisition years are added context (public-record approximations) so the lanes can be drawn; era labels are editorial. The 0–12 month roadmap phases are from section 9.2 of the analysis (relative to 20 November 2026).